Larry: 20 Dollar Bills and Factor Persistence

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Larry: 20 Dollar Bills and Factor Persistence

Post by Call_Me_Op »

I was listening this morning to a radio show from 2011 where Larry Swedroe provided an analogy to efficient markets. Basically, he said that if it were known that there were a lot of 20 dollar bills lying on the ground in a certain area, people would rush to that area to scoop them up. His point was that it n general you will not find many 20 dollar bills just lying around. And thus it is not worth the effort to go looking for them. In other words, the market is quite efficient.

I got to thinking about small-cap value - in relation to Larry's analogy. It is now widely known that small-cap value has significantly outperformed. Utilizing Larry's analogy, it seems logical that many more people will invest in small-cap value in the future, reducing long-term returns. If Larry sees this post, I would be interested in his rebuttal. Comments from others are (of course) welcome.
Last edited by Call_Me_Op on Sun Jul 13, 2014 7:00 am, edited 1 time in total.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
kenner
Posts: 3128
Joined: Sat Mar 01, 2008 7:45 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by kenner »

Reward follows risk, eventually. Assuming you are properly diversified.
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by Call_Me_Op »

kenner wrote:Reward follows risk, eventually. Assuming you are properly diversified.
Sure, but by "risk", you are talking about standard deviation of stock returns. It is quite conceivable that the increase in investment dollars could reduce both returns and risk. After all, why have value stocks been risky? They have been out-of-favor (there have been low expectations for profit growth) and the underlying companies occasionally surprise to the upside. It is more difficult to argue that they will be out-of-favor going forward since stock prices may be driven more by expectations of a SV premium as opposed to being based on analyses of underlying fundamentals.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by larryswedroe »

Call me op
What we should expect to see given information has been now well known for at least 20 years now is that the premiums might shrink if risk based and be more likely to do so if it is behavioral. Since these stocks are expensive to short in many cases, and there are limits to arbitrage as well as costs, and you have behavioral issues, at best I would think might shrink. That is what has happened to basically most of the premium--that is what Claude Erb found ---I wrote series of articles at SeekingAlpha on that starting with this one http://seekingalpha.com/article/2212683 ... overgrazed

But he had not found that the small value premium had shrunk. Now by now that is possible, one has to look at the spread of btm (etc) between small value and small growth----Davis of DFA showed that is predictor of premium---to see if that spread has shrunk.

As to whether it will or not my crystal ball is cloudy as usual---but if you think that will happen you should load up on small value now to capture the gains from that happening--as valuations of SV rise
Larry
steve_14
Posts: 1507
Joined: Wed Jun 20, 2012 12:05 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by steve_14 »

Call_Me_Op wrote:It is now widely known that small-cap value has significantly outperformed. Utilizing Larry's analogy, it seems logical that many more people will invest in small-cap value in the future, reducing long-term returns. If Larry sees this post, I would be interested in his rebuttal. Comments from others are (of course) welcome.
Considering small and value factors separately, and looking at real fund results, not reconstructed backtests, will probably go a long way toward answering your question. There's been no free lunch or extra return in value stocks - compare your favorite value/growth pairing and you'll see that. Each has had periods of outperformance, as random variation would dictate. Nor has there been any free lunch in small caps, which have had periods of violent underperformance (80s and 90s), and outperformance (last 15 years).

Our randomly walking market is working efficiently, just as it should be.
heyyou
Posts: 4461
Joined: Tue Feb 20, 2007 3:58 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by heyyou »

Thanks Steve, you are more articulate than I am, but here is mine anyway.

Factors by definition are durable but a consistent return higher than other portions of the market is not to be expected. That depends on the both the return of the targeted slice and the returns of the rest of the market slices.

The lack of patterns in a Callan Periodic Table shows how variable the returns of sub-asset classes are. Ignore the bonds, and there is still no consistent pattern of the superiority of SCV to Large Caps in that table (large caps as a proxy for total market). Over long periods, SCV is expected to out perform Large Caps on average, but only by having large swings between under-performing then over-performing.

The fickleness of investors is a constant, but their favorite investments change often, including those based on factors.

My opinion is that there is little certainty anywhere in the markets. If there was, someone clever would figure out how to take advantage of it. Seems like the OP is asking Larry to somehow defend the immediate future performance of factors, which is not a fair question.

Here is one answer. Buy a balanced, diversified but tilted portfolio, hold for the next thirty years, then compare with a no-tilt portfolio. To remove some of the facetiousness, I bought my first index fund shares, 30-31 years ago. It has been a bumpy ride, but the results are wonderful.
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by Call_Me_Op »

heyyou wrote: The lack of patterns in a Callan Periodic Table shows how variable the returns of sub-asset classes are. Ignore the bonds, and there is still no consistent pattern of the superiority of SCV to Large Caps in that table (large caps as a proxy for total market). Over long periods, SCV is expected to out perform Large Caps on average, but only by having large swings between under-performing then over-performing.
I don't think anyone would argue that factors mean consistent out-performance. At least historically, they have corresponded to out-performance when performance is viewed over very long time periods.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
cowboysFan
Posts: 290
Joined: Sat Jan 25, 2014 12:46 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by cowboysFan »

Call_Me_Op wrote:
kenner wrote:Reward follows risk, eventually. Assuming you are properly diversified.
Sure, but by "risk", you are talking about standard deviation of stock returns.
The Fama school of thought says that standard deviation is not risk and therefore Sharpe ratios do not represent risk adjusted returns. This point has been brought up many times in past discussions.
cowboysFan
Posts: 290
Joined: Sat Jan 25, 2014 12:46 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by cowboysFan »

larryswedroe wrote: Since these stocks are expensive to short in many cases, and there are limits to arbitrage as well as costs, and you have behavioral issues, at best I would think might shrink.
Isn't the August 2007 hedge fund meltdown believed to have occurred because all the major hedge fund shops had been applying the same long/short factor based strategies, so aren't factor based strategies something the hedge fund industry has been aware of and using for years now?
steve_14
Posts: 1507
Joined: Wed Jun 20, 2012 12:05 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by steve_14 »

heyyou wrote:Thanks Steve, you are more articulate than I am, but here is mine anyway.

Factors by definition are durable but a consistent return higher than other portions of the market is not to be expected. That depends on the both the return of the targeted slice and the returns of the rest of the market slices.

The lack of patterns in a Callan Periodic Table shows how variable the returns of sub-asset classes are. Ignore the bonds, and there is still no consistent pattern of the superiority of SCV to Large Caps in that table (large caps as a proxy for total market). Over long periods, SCV is expected to out perform Large Caps on average, but only by having large swings between under-performing then over-performing.

The fickleness of investors is a constant, but their favorite investments change often, including those based on factors.

My opinion is that there is little certainty anywhere in the markets. If there was, someone clever would figure out how to take advantage of it. Seems like the OP is asking Larry to somehow defend the immediate future performance of factors, which is not a fair question.

Here is one answer. Buy a balanced, diversified but tilted portfolio, hold for the next thirty years, then compare with a no-tilt portfolio. To remove some of the facetiousness, I bought my first index fund shares, 30-31 years ago. It has been a bumpy ride, but the results are wonderful.
That sounds about right to me - unfortunately, people have a hard time understanding randomness, or fickleness, or random variation. But patterns come naturally. That's why the #1 sales tool of financial advisors is a chart showing the historical outperformance of [active mutual fund/strategy/stock/private REIT/whatever is being sold]. That's all it takes - continued outperformance is now assumed by the client.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by larryswedroe »

Considering small and value factors separately, and looking at real fund results, not reconstructed backtests, will probably go a long way toward answering your question. There's been no free lunch or extra return in value stocks - compare your favorite value/growth pairing and you'll see that.
This statement is totally without merit.

Here's just one example

Just look at the returns of DFA SV since inception, which comes very close to matching the returns of its benchmark index and it way outperforms small growth funds. If it was a horse race they would have shot small growth long ago. 4/93-6/14 DFA SV fund returned 13.04 vs 7.74 for the Russell small growth index which has no costs.


I would add this about this quotation
Our randomly walking market is working efficiently, just as it should be.

Markets can work randomly but premiums persist---in other words there is a positive expected return but it isn't consistent. They move up randomly with about an 8% annual risk premium. But market is highly efficient.
The same is true for size and value and MOM, etc. Their returns aren't random, If they were there would be a zero return to them



Larry



Larry
Last edited by larryswedroe on Sun Jul 13, 2014 8:54 pm, edited 1 time in total.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by larryswedroe »

cowboysfan
The premise of your statement isn't correct .
the August 2007 hedge fund meltdown believed to have occurred because all the major hedge fund shops had been applying the same long/short factor based strategies
But the latter part is true--that they are well aware of the research. Costs kill them. And active strategies and their costs, not factor exposures

Also I don't know where you get the idea that the "Fama school" of thought doesn't think SD is risk and that's why SR isn't good measure

Personally I don't know what Fama would say but would guess he would say SD is one type of risk, as it is used in the studies he does. Adding SR meaningful ONLY if returns are normally distributed

Larry
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by larryswedroe »

Call me Op
I don't think anyone would argue that factors mean consistent out-performance. At least historically, they have corresponded to out-performance when performance is viewed over very long time periods.
Yes, that of course is true, but I would use the word PERSISTENT, which is different than consistent. And they get called factors because the evidence is both persistent and pervasive and statistically significant
That's what the research shows very clearly. Anyone interested should read books like Expected Returns, The Physics of Wall Street and Successful Investing is a Process and the Success Equation.

Larry
cowboysFan
Posts: 290
Joined: Sat Jan 25, 2014 12:46 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by cowboysFan »

larryswedroe wrote:cowboysfan
The premise of your statement isn't correct .
the August 2007 hedge fund meltdown believed to have occurred because all the major hedge fund shops had been applying the same long/short factor based strategies
But the latter part is true--that they are well aware of the research. Costs kill them. And active strategies and their costs, not factor exposures

Larry
I think you're being glib. Management costs(2 and 20) can explain why hedge funds might under-perform as a group, but those costs were an issue decades before August 2007 and in the years after 2007. They certainly don't explain why all the quant funds suffered major losses the same month until Goldman Sachs injected billions in capital and stopped the bleeding.

The bigger point I was trying to make was that SV and momentum strategies have been getting more popular over the past two decades as a lot of the quant hedge funds have been deploying a lot of capital to capture those premiums.
Last edited by cowboysFan on Sun Jul 13, 2014 11:07 pm, edited 1 time in total.
cowboysFan
Posts: 290
Joined: Sat Jan 25, 2014 12:46 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by cowboysFan »

larryswedroe wrote: Also I don't know where you get the idea that the "Fama school" of thought doesn't think SD is risk and that's why SR isn't good measure

Personally I don't know what Fama would say but would guess he would say SD is one type of risk, as it is used in the studies he does. Adding SR meaningful ONLY if returns are normally distributed

Larry
Perhaps, I should have said Fama believes SD is only one of several types of risks. In past interviews, he's referred to small and value as really risk premiums for the types of risks that can't be captured by beta. If you think there are three or four or five types of risk, then the risk-adjusted out-performance disappears if you use a more comprehensive definition of risk than just a Sharpe ratio. Since then he's moved on to a four factor or five factor model or maybe it's six by now, but I doubt his basic beliefs have changes much.
acegolfer
Posts: 3029
Joined: Tue Aug 25, 2009 9:40 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by acegolfer »

OP,

Is this your argument? Since factor SV premiums have been persistent, the market is inefficient? (Sorry, if I have misinterpreted.)

If yes, then what about market premium? If market is efficient, are you suggesting the market premium should not be persistent because there's no free lunch?
steve_14
Posts: 1507
Joined: Wed Jun 20, 2012 12:05 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by steve_14 »

larryswedroe wrote:
Considering small and value factors separately, and looking at real fund results, not reconstructed backtests, will probably go a long way toward answering your question. There's been no free lunch or extra return in value stocks - compare your favorite value/growth pairing and you'll see that.
This statement is totally without merit.

Here's just one example

Just look at the returns of DFA SV since inception, which comes very close to matching the returns of its benchmark index and it way outperforms small growth funds. If it was a horse race they would have shot small growth long ago. 4/93-6/14 DFA SV fund returned 13.04 vs 7.74 for the Russell small growth index which has no costs.
I'd say that's not one example, but the only example - for a few years (pre Reg FD I might add) the DFA SV fund, like many non-index funds, had a period of outperformance. And if that had been sustainable, you might have an argument for a persistent premium. But, as I've shown repeatedly, it was not.

Looking back at the Diehards forum archives, "How to I get access to DFA" was a question that came up repeatedly. Not so much anymore.
kenner
Posts: 3128
Joined: Sat Mar 01, 2008 7:45 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by kenner »

Call_Me_Op wrote:
kenner wrote:Reward follows risk, eventually. Assuming you are properly diversified.
Sure, but by "risk", you are talking about standard deviation of stock returns.

No, I'm not. Standard deviation is only one measure of risk.


It is quite conceivable that the increase in investment dollars could reduce both returns and risk. After all, why have value stocks been risky? They have been out-of-favor (there have been low expectations for profit growth) and the underlying companies occasionally surprise to the upside. It is more difficult to argue that they will be out-of-favor going forward since stock prices may be driven more by expectations of a SV premium as opposed to being based on analyses of underlying fundamentals.
kenner
Posts: 3128
Joined: Sat Mar 01, 2008 7:45 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by kenner »

cowboysFan wrote: The Fama school of thought says that standard deviation is not risk and therefore Sharpe ratios do not represent risk adjusted returns. This point has been brought up many times in past discussions.
So are you declaring for all eternity that the "Fama School" is 100% correct, no room for debate?
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by Call_Me_Op »

acegolfer wrote:OP,

Is this your argument? Since factor SV premiums have been persistent, the market is inefficient? (Sorry, if I have misinterpreted.)

If yes, then what about market premium? If market is efficient, are you suggesting the market premium should not be persistent because there's no free lunch?
I am just trying to reconcile Larry's argument regarding 20 dollar bills. But perhaps this can be couched as you suggest - as a question of whether an SV premium can exist in an efficient market. And I suppose you are right - an extension of this question is can beta exist in an efficient world. So the answer to both must be "yes" - provided that risk and return are commensurate.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
richard
Posts: 7961
Joined: Tue Feb 20, 2007 2:38 pm
Contact:

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by richard »

Call_Me_Op wrote:
acegolfer wrote:OP,

Is this your argument? Since factor SV premiums have been persistent, the market is inefficient? (Sorry, if I have misinterpreted.)

If yes, then what about market premium? If market is efficient, are you suggesting the market premium should not be persistent because there's no free lunch?
I am just trying to reconcile Larry's argument regarding 20 dollar bills. But perhaps this can be couched as you suggest - as a question of whether an SV premium can exist in an efficient market. And I suppose you are right - an extension of this question is can beta exist in an efficient world. So the answer to both must be "yes" - provided that risk and return are commensurate.
Beta is clearly a risk - stocks are riskier than bonds. If nothing else, bonds are mandated payments while stocks are the residual value.

SV is not clearly a risk. It may be a proxy for some underlying risk, such as distress, but is not itself a risk.

Therefore, the answers to both questions don't have to be the same. They might be or they might not be.

The efficient market claim is not that risk and return are commensurate. If more risk necessarily led to more return, risk wouldn't be risky. The claim is that more risk increases your chances of a higher return (else why take on more risk).
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by Call_Me_Op »

richard wrote:
Call_Me_Op wrote:
acegolfer wrote:OP,

Is this your argument? Since factor SV premiums have been persistent, the market is inefficient? (Sorry, if I have misinterpreted.)

If yes, then what about market premium? If market is efficient, are you suggesting the market premium should not be persistent because there's no free lunch?
I am just trying to reconcile Larry's argument regarding 20 dollar bills. But perhaps this can be couched as you suggest - as a question of whether an SV premium can exist in an efficient market. And I suppose you are right - an extension of this question is can beta exist in an efficient world. So the answer to both must be "yes" - provided that risk and return are commensurate.
Beta is clearly a risk - stocks are riskier than bonds. If nothing else, bonds are mandated payments while stocks are the residual value.

SV is not clearly a risk. It may be a proxy for some underlying risk, such as distress, but is not itself a risk.

Therefore, the answers to both questions don't have to be the same. They might be or they might not be.

The efficient market claim is not that risk and return are commensurate. If more risk necessarily led to more return, risk wouldn't be risky. The claim is that more risk increases your chances of a higher return (else why take on more risk).
Looking forward, risk is commensurate with (or more correctly, proportional to) expected return.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
acegolfer
Posts: 3029
Joined: Tue Aug 25, 2009 9:40 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by acegolfer »

Call_Me_Op wrote: Looking forward, risk is commensurate with (or more correctly, proportional to) expected return.
Asset pricing models (CAPM, FF, APT) have linear functions. E(R) - Rf = b1 * RP1 + b2 * RP2 + b3 * RP3 + .....
richard
Posts: 7961
Joined: Tue Feb 20, 2007 2:38 pm
Contact:

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by richard »

Call_Me_Op wrote:
richard wrote:
Call_Me_Op wrote:
acegolfer wrote:OP,

Is this your argument? Since factor SV premiums have been persistent, the market is inefficient? (Sorry, if I have misinterpreted.)

If yes, then what about market premium? If market is efficient, are you suggesting the market premium should not be persistent because there's no free lunch?
I am just trying to reconcile Larry's argument regarding 20 dollar bills. But perhaps this can be couched as you suggest - as a question of whether an SV premium can exist in an efficient market. And I suppose you are right - an extension of this question is can beta exist in an efficient world. So the answer to both must be "yes" - provided that risk and return are commensurate (emphasis added).
Beta is clearly a risk - stocks are riskier than bonds. If nothing else, bonds are mandated payments while stocks are the residual value.

SV is not clearly a risk. It may be a proxy for some underlying risk, such as distress, but is not itself a risk.

Therefore, the answers to both questions don't have to be the same. They might be or they might not be.

The efficient market claim is not that risk and return are commensurate. If more risk necessarily led to more return, risk wouldn't be risky. The claim is that more risk increases your chances of a higher return (else why take on more risk).
Looking forward, risk is commensurate with (or more correctly, proportional to) expected return.
Expected returns are not the same as returns - I underlined the relevant portion of your earlier post.
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by Call_Me_Op »

richard wrote: Expected returns are not the same as returns - I underlined the relevant portion of your earlier post.
I am aware of that - which is why I clarified in my follow-up.

If risk is taken to mean standard deviation of returns (as common per academics), then we can specify neither risk or return in the future with certainty. So both must be specified statistically. (However, we can deterministically specify both looking backward.) So you can modify my statement to read "...provided that our expectation of risk and return are commensurate."
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by larryswedroe »

cowboysfan
Wasn't being glib at all. The value factor and small factors have been known for decades and hedge funds have been underperforming since late 90s. And value small value has outperformed since then. The 2007 crisis had very little to do with hedge funds.

Now the amount of assets has increased dramatically in hedge funds in general, and as result of that plus in general more money flowing into institutional "smart" investors (vs retail "dumb" money) the harder it is for hedge funds to win (fewer victims to exploit) and also the likelihood that factor premiums might shrink--which is exactly what has happened--as Erb found. But small value has not shrunk, at least not yet. Doesn't mean it won't. But Erb also found that in the asset classes that are the least liquid and hardest to arb the premiums have shrunk the least or not shrunk

But if you believe that the premium will shrink than you should load up on exposure to it know, because the way it shrinks is to see valuations rise. So you want to be the first in.
Also if you believe it's risk, not free lunch, than it might not shrink at all.

Larry
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by larryswedroe »

Steve
Sorry but so much of what you write, as has been the case in this thread, is just wrong. You can have your opinions but you state things as facts that are not true.

You data mine extensively to find some subperiod and claim that as proof when no one states that factors cannot or will not underperform for even very long periods or even forever. As I have stated there is a non zero chance that any factor will underperform no matter the horizon. What we should look for is data that is persistent and pervasive, not single out subperiods, especially very short ones that you use. And we should use the longest periods we have.

As to this statement,
I'd say that's not one example, but the only example - for a few years (pre Reg FD I might add) the DFA SV fund, like many non-index funds, had a period of outperformance. And if that had been sustainable, you might have an argument for a persistent premium. But, as I've shown repeatedly, it was not.


This too is a false statement, or misleading at best. What we had was period where size/value underperformed (just as beta can) and DFA underperforms Vanguard because it has more exposure to the factor that is LIKELY TO OUTPERFORM. That was reasonably short period and now the last 5 years DFA has outperformed (so clearly Ref FD has nothing to do with this issue). And most importantly for last 15 years it has way outperformed,

Note that this has nothing to do with DFA either as one can use other funds that have more exposure to the factors and find the same thing if they also execute well.

Larry
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by larryswedroe »

acegolfer
Premiums do not mean market is inefficient. Beta is a factor and no one argues that because there is beta that markets are inefficient.
Now if you have premiums without a risk story that is logical you can then say you think market is inefficient. MOM falls into that category IMO, and it's hard to make an economic case for profitability premium without believing market inefficient (though I can do that, has to do with duration) .

But then you have to decide if once published will anomaly persist---or will limits to arbitrage and fear and cost of leverage/margin (as well as the human behavior that leads to the anomaly persisting) will prevent the factor from being arbed away

Larry
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by Browser »

Larry - since value, small, or small value can underperform for years and will do so in the future, and because - as you state - there's no guarantee these premiums will persist in the future - isn't there a prospective dilemma for investors if (when) their factor-based investment allocations stay down for several years? How will they know whether this is cyclical risk or the premium has left the building? How should investors think about this possible scenario if they decide to commit a significant part of their portfolio to these types of investments?
We don't know where we are, or where we're going -- but we're making good time.
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by Call_Me_Op »

Browser wrote:Larry - since value, small, or small value can underperform for years and will do so in the future, and because - as you state - there's no guarantee these premiums will persist in the future - isn't there a prospective dilemma for investors if (when) their factor-based investment allocations stay down for several years? How will they know whether this is cyclical risk or the premium has left the building? How should investors think about this possible scenario if they decide to commit a significant part of their portfolio to these types of investments?
You will not know (whether the risk is cyclical or the premium has "left the building"). That is the nature of uncertainty. My recommendation is not to rely upon persistence of factor premiums - although you can hope for it and invest accordingly.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by larryswedroe »

Browser

That's the real problem for many if not most investors---they really don't have the knowledge IMO needed to be successful so they confuse strategy with outcome and fall prey to recency and tracking error regret. That's why one should never invest in anything they don't fully understand the nature of the risks. IMO it's one of the biggest value adds of a good advisor, educating investors so they avoid those mistakes and then holding their hands in periods of underperformance that inevitably occur so they stay disciplined. We have gone through some long periods of underperformance and yet our client turnover rate is about 1%, and that is IMO because of the large amount of UPFRONT education we do---battles are won in the preparation stage, not on battlefield. Forewarned is forearmed

Hope that is helpful

Larry
acegolfer
Posts: 3029
Joined: Tue Aug 25, 2009 9:40 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by acegolfer »

larryswedroe wrote:acegolfer
Premiums do not mean market is inefficient. Beta is a factor and no one argues that because there is beta that markets are inefficient.
Now if you have premiums without a risk story that is logical you can then say you think market is inefficient. MOM falls into that category IMO, and it's hard to make an economic case for profitability premium without believing market inefficient (though I can do that, has to do with duration) .

But then you have to decide if once published will anomaly persist---or will limits to arbitrage and fear and cost of leverage/margin (as well as the human behavior that leads to the anomaly persisting) will prevent the factor from being arbed away

Larry
I totally agree with you. I was trying to show OP that his argument was incomplete by showing the fallacy in his logic. There's no "risk" in OP.
acegolfer
Posts: 3029
Joined: Tue Aug 25, 2009 9:40 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by acegolfer »

Another note:

Testing market efficiency is a joint hypothesis test and it's never possible to disprove market is efficient. For example, even if other non-risk premium persists, that doesn't mean market is inefficient. It could be due to incomplete model. In addition, some portfolios such as SMB, HML, MOM, PROF may not really be risk factors but they could be factor mimicking portfolios without rational risk theory.

Here's a related wiki: http://en.wikipedia.org/wiki/Efficient- ... is_problem
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by Browser »

William Sharpe describes an exercise he performs with his students. He asks them to set up a spreadsheet based on price data with a known, actual risk premium of 5.5%. He then asks them to generate 1000 series of random returns from that distribution over 50-year periods and compute the observed risk premiums. You observe risk premiums over a very wide range, from zero to twenty percent over 50 years, even when there is a known, fixed risk premium of 5.5% underlying the data. As he points out with this example, anyone who thinks that empirical data show, with any degree of precision, that the risk premium for small, value, or small value stocks is any different from the premium for the total market is "just kidding himself." If Sharpe is correct, it seems like folly to invest in a tiny corner of the total market such as SV and persist in holding onto this investment for what could turn out to be a really long period of market underperformance even if the premium is real and even if distributions never change. Myself, I would find it next to impossible to overcome "Sharpe's Skepticism". It is hard to regard such a strategy outside the confines of a "gamble" that might be worth a small fraction of one's stake but not much more.
We don't know where we are, or where we're going -- but we're making good time.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by larryswedroe »

Browser
How about flipping it the other way, if you cannot distinguish them why would you want to have all your eggs in one factor basket (beta)?
Given that the prudent strategy in the presence of uncertainty is diversification of risks why concentrate when you have no more confidence in the beta premium than the others showing up? And even have have more confidence in beta (not sure why that would be the case since SV outperforms market as much as beta outperforms in terms of persistence) is it enough confidence to hold all your eggs in beta basket?
Just asking
Larry
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by Call_Me_Op »

acegolfer wrote:
larryswedroe wrote:acegolfer
Premiums do not mean market is inefficient. Beta is a factor and no one argues that because there is beta that markets are inefficient.
Now if you have premiums without a risk story that is logical you can then say you think market is inefficient. MOM falls into that category IMO, and it's hard to make an economic case for profitability premium without believing market inefficient (though I can do that, has to do with duration) .

But then you have to decide if once published will anomaly persist---or will limits to arbitrage and fear and cost of leverage/margin (as well as the human behavior that leads to the anomaly persisting) will prevent the factor from being arbed away

Larry
I totally agree with you. I was trying to show OP that his argument was incomplete by showing the fallacy in his logic. There's no "risk" in OP.
Say what? To what argument do you refer? I do not recall presenting an argument - I thought I was asking a question.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
acegolfer
Posts: 3029
Joined: Tue Aug 25, 2009 9:40 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by acegolfer »

@OP,

This is the part that I was referring to.
Utilizing Larry's analogy, it seems logical that many more people will invest in small-cap value in the future, reducing long-term returns.
If that's not your argument, my apologies. Throughout the entire thread, I thought you were using Larry's $20 bill to argue that SC premiums should disappear. Clearly, I misinterpreted the OP.
User avatar
packer16
Posts: 1488
Joined: Sat Jan 04, 2014 1:28 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by packer16 »

Browser wrote:William Sharpe describes an exercise he performs with his students. He asks them to set up a spreadsheet based on price data with a known, actual risk premium of 5.5%. He then asks them to generate 1000 series of random returns from that distribution over 50-year periods and compute the observed risk premiums. You observe risk premiums over a very wide range, from zero to twenty percent over 50 years, even when there is a known, fixed risk premium of 5.5% underlying the data. As he points out with this example, anyone who thinks that empirical data show, with any degree of precision, that the risk premium for small, value, or small value stocks is any different from the premium for the total market is "just kidding himself." If Sharpe is correct, it seems like folly to invest in a tiny corner of the total market such as SV and persist in holding onto this investment for what could turn out to be a really long period of market underperformance even if the premium is real and even if distributions never change. Myself, I would find it next to impossible to overcome "Sharpe's Skepticism". It is hard to regard such a strategy outside the confines of a "gamble" that might be worth a small fraction of one's stake but not much more.
This is an example that I am very skeptical of academic research. This exercise is having a theory, running an experiment finding a result that is no where near what the historical data shows and somehow saying this "proves" that the data cannot "prove" a premium exists. These guys are using scientific tests to try to find trends in economic data. Using that level of required precision does this imply that the ERP does not exist? This is physics envy at its finest.

Packer
Buy cheap and something good might happen
User avatar
telemark
Posts: 3389
Joined: Sat Aug 11, 2012 6:35 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by telemark »

Questions like this make me wish I'd spent less time studying abstract algebra and computability theory and more time on plain old statistics. Without the necessary background to check Larry's work, I can only say that one should always be careful of arguments that say "I have a model that says this can't happen, therefore it will never happen." The real world has a way of being more complicated than our attempts to model it. If you want to say "the difference is small enough that I'm not going to worry about it", then you're on ground that any engineer can understand, but that's a very different position.

Oh, and the $20 bill thing is not a formal statement of the EMH, just an example to give students a quick idea of how it's supposed to work. You shouldn't take it as an unbreakable law.
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by Call_Me_Op »

acegolfer wrote:@OP,

This is the part that I was referring to.
Utilizing Larry's analogy, it seems logical that many more people will invest in small-cap value in the future, reducing long-term returns.
If that's not your argument, my apologies. Throughout the entire thread, I thought you were using Larry's $20 bill to argue that SC premiums should disappear. Clearly, I misinterpreted the OP.
It was posed as a question. I was asking Larry (in effect) that if what he says about 20 dollars bills is true, why does this not extend to a segment of the market believed to have higher reward than other segments - just as a segment of the ground populated by 20 dollar bills is perceived to offer a higher reward than other segments of the ground. I think the answer is "we don't know whether or not the SV premium will go the way of the 20 dollar bills."
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
steve_14
Posts: 1507
Joined: Wed Jun 20, 2012 12:05 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by steve_14 »

Call_Me_Op wrote:I think the answer is "we don't know whether or not the SV premium will go the way of the 20 dollar bills."
I think that's a safe assumption. There's going to be a "premium" somewhere over the next 30 years, but if anyone knew where it would be, it would be instantly traded away. Of course, it's the job of the investment industry to convince you up is down, red is green, and persistent free lunches are there for the taking. TSM funds with .05% ERs don't generate much in the way of profits.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by larryswedroe »

Call me op
I would say it differently
I would say that while the odds of SV premium disappearing are non zero, the evidence suggests that the odds of that happening are about the same as the odds of the beta premium disappearing


Telemark

First I would never say what you put in quotations "I have a model that says this can't happen, therefore it will never happen." In fact I've written often that just because something hasn't happened doesn't mean it cannot or will not. I've also written often about not confusing the unlikely as impossible nor the highly likely as certain. And finally the evidence I present generally isn't based on "my work" but on the evidence from peer reviewed academic journals (where qualified people have reviewed the work).


Larry
User avatar
telemark
Posts: 3389
Joined: Sat Aug 11, 2012 6:35 am

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by telemark »

larryswedroe wrote: Telemark

First I would never say what you put in quotations "I have a model that says this can't happen, therefore it will never happen." In fact I've written often that just because something hasn't happened doesn't mean it cannot or will not. I've also written often about not confusing the unlikely as impossible nor the highly likely as certain. And finally the evidence I present generally isn't based on "my work" but on the evidence from peer reviewed academic journals (where qualified people have reviewed the work).
Larry
It was not your position that I was characterizing in that possibly unfair way, but rather the one that seems to say "the EMH says that anomalies can't persist, so they don't, period." Sorry for the confusion. I also did not intend any misattribution and appreciate the correction there.
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by Browser »

larryswedroe wrote:Browser
How about flipping it the other way, if you cannot distinguish them why would you want to have all your eggs in one factor basket (beta)?
Given that the prudent strategy in the presence of uncertainty is diversification of risks why concentrate when you have no more confidence in the beta premium than the others showing up? And even have have more confidence in beta (not sure why that would be the case since SV outperforms market as much as beta outperforms in terms of persistence) is it enough confidence to hold all your eggs in beta basket?
Just asking
Larry
The question was asked of Sharpe. His short answer was that, yes there is plenty of research that shows that riskier diversified equity portfolios have outperformed less risky portfolios (his view is that the most efficient basket for your equity eggs is the total market, and when you allocate your eggs differently you may be taking on more risks, and you are also assuming higher costs). So, it might make sense to do this but also to be skeptical about putting too much money into active management and chasing anomalies. You have to ask yourself why, in a hugely competitive market, one should expect a particular anomaly to happen in the future with sufficient certainty to justify putting much money into it. As one can see, his view is that anomalies can well be the result of not much more than random historical data, even over period of 50 years or longer.
We don't know where we are, or where we're going -- but we're making good time.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by larryswedroe »

Browser
First I don't think SV is an anomaly, at best the premium may have been too large for the extra risks, but even that's hard to argue since the SD of SV is about 2x that of the market. I would add if there are limits to arbitrage and high costs/risks of margin/shorting than anomalies can and do persist (there are many proving that the markets are not perfectly efficient).

Second, as I show in Reducing the Risk of Black Swans, at least historically a low beta and high tilt portfolio has been much more efficient than a market like portfolio--not even close---And I would note that we have been using the strategy for about 20 years now so it's not like we are looking backwards solely. Investors who used the strategy clearly benefited and certainly slept a lot better through the bear markets of 2000-02 and late 2007- early 2009. Now whether it will be more efficient going forward I don't know, though I believe so and present the reasons why in the book--the diversification benefits of a multi-factor approach and how the risks mix much better than a TSM portfolio if use the highest quality bonds

Finally, never get me to argue that a TSM portfolio isn't a perfectly reasonable one. I just don't think it's the most efficient. But that's my opinion of course

Hope that is helpful

Larry
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by Browser »

Finally, never get me to argue that a TSM portfolio isn't a perfectly reasonable one. I just don't think it's the most efficient. But that's my opinion of course
As they say, a difference of opinion makes a market. Plenty of credible opinion on the other side of the argument. How it could come to pass that more than a tiny minority of investors could be convinced about SCV and the "premium" not be arbitraged away is beyond me. If I really believed it, I'd try to keep it a secret instead of publishing articles and books touting it. :) So, let's fervently hope that more people believe what I believe than what DFA believes and they don't pay any attention to LS. :wink:
We don't know where we are, or where we're going -- but we're making good time.
cowboysFan
Posts: 290
Joined: Sat Jan 25, 2014 12:46 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by cowboysFan »

kenner wrote:
cowboysFan wrote: The Fama school of thought says that standard deviation is not risk and therefore Sharpe ratios do not represent risk adjusted returns. This point has been brought up many times in past discussions.
So are you declaring for all eternity that the "Fama School" is 100% correct, no room for debate?
Philosophically, it comes down to if you think markets are reasonably efficient. I think if there ever was a risk free and widely known way to beat the market by 1% a year, soon so many institutional investors would be following that strategy that the out-performance would disappear. If Larry's right and the premium for SV is really a behavioral anomaly, it's probably well on it's way to being arbitraged out of existence.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by larryswedroe »

Browser
You seem to keep missing the point that if it's risk that leads to the premium doesn't matter how many people are aware of it it won't go away----in other words it's not higher risk adjusted returns (at least in isolation). No different than everyone aware of BETA premium but people don't fully invest in stocks.
Note either way SV investors win. They win in short term by valuations of SV rising to eliminate the premium, or they win in long term by earning the EXPECTED premium, not guaranteed of course. So if you think it will be arbed away because it isn't risk you should be loading up on SV. If you think it's risk then you have different question, or set of questions, to answer.
Larry
User avatar
packer16
Posts: 1488
Joined: Sat Jan 04, 2014 1:28 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by packer16 »

One aspect of SCV or maybe micro-cap value that is hard to arb away is liquidity. If you look at many of these micro-caps there is just not enough volume for most institutional investors to be interested. For example, if you have a $100 million fund and you need lets say 50 stocks to have a diversified portfolio then each position has to be $2 million. Lets say you don't want to own more than 1% of a company then that limits you to firms with market caps greater than $200 million. There are alot of stocks below $200 million out there with not many folks looking at them because the economic incentive is not there. In addition, there is a 5% limit on total ownership of any fund. If you charge 1% fee you generate $1 million before expenses. However, to track and stay abreast of a 50 stock portfolio it is probably going to require more than you and what about trading, marketing and back office. This is just to start a fund assuming your not going to be providing financial planning which will cost more. Funds will for the most part not touch these smaller stocks due to redemption risk also. If you also go to foreign restricted markets (like S. Korea) it is even worse in terms of arb. Also the situation we are speaking of here (mispriced securities) cannot be programmed into a computer. If you have every looked at outputs of stock screens you understand why, many of these firms that appear cheap are not if the data is adjusted correctly. I just think some folks have a warped view of the less than $200 million space is like the megacap multi-billion space and the spaces are very different. I think the consolidation trend in money management is just making the differences worse.

Packer
Buy cheap and something good might happen
cowboysFan
Posts: 290
Joined: Sat Jan 25, 2014 12:46 pm

Re: Larry: 20 Dollar Bills and Factor Persistence

Post by cowboysFan »

packer16 wrote:One aspect of SCV or maybe micro-cap value that is hard to arb away is liquidity. If you look at many of these micro-caps there is just not enough volume for most institutional investors to be interested. For example, if you have a $100 million fund and you need lets say 50 stocks to have a diversified portfolio then each position has to be $2 million. Lets say you don't want to own more than 1% of a company then that limits you to firms with market caps greater than $200 million. There are alot of stocks below $200 million out there with not many folks looking at them because the economic incentive is not there.
Packer
The flip side of that is Rick Ferri has argued a true index fund in the 10th decile crsp doesn't exist and even BRISX has a large tracking error with respect to its benchmark. If you wanted to try to roll your own index, as opposed to picking individual stocks, with say at least 200 stocks and don't get killed with trading costs and bid-ask spreads, the taxes alone would be hell.
Post Reply