Better growth? Or on how the value premium ends

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
stlutz
Posts: 5427
Joined: Fri Jan 02, 2009 1:08 am

Re: Better growth? Or on how the value premium ends

Post by stlutz » Fri Aug 19, 2016 6:46 pm

You can also see that the Value premium is smaller for Large Stocks than for Small Stocks. From what I have been reading, it seems that the Value premium for Small Stocks is unchanged.
Not really. In the smallcap space, much of the "value premium" is really a "growth penalty", and this penalty is most significant in stocks that aren't in the indexes or in institutional portfolios for various reasons. Just by starting out with the "base screen" of companies that index providers or fund managers like DFA use eliminates much of the growth penalty.

See this thread for further discussion: viewtopic.php?t=157764

User avatar
nedsaid
Posts: 12668
Joined: Fri Nov 23, 2012 12:33 pm

Re: Better growth? Or on how the value premium ends

Post by nedsaid » Fri Aug 19, 2016 7:15 pm

stlutz wrote:
You can also see that the Value premium is smaller for Large Stocks than for Small Stocks. From what I have been reading, it seems that the Value premium for Small Stocks is unchanged.
Not really. In the smallcap space, much of the "value premium" is really a "growth penalty", and this penalty is most significant in stocks that aren't in the indexes or in institutional portfolios for various reasons. Just by starting out with the "base screen" of companies that index providers or fund managers like DFA use eliminates much of the growth penalty.

See this thread for further discussion: viewtopic.php?t=157764
Is the "growth penalty" the lottery tickets that Larry talks about?
A fool and his money are good for business.

stlutz
Posts: 5427
Joined: Fri Jan 02, 2009 1:08 am

Re: Better growth? Or on how the value premium ends

Post by stlutz » Fri Aug 19, 2016 7:32 pm

A few things which this thread highlights that I'd like to point out:

--From the perspective of stock picking, "risk" vs "behavioral" explanations of historical anomalies are a distinction without a difference. In political economics, the efficient market and behavioral schools have huge disagreements that matter. In this thread, Larry talks about "risk" while nedsaid talks about "behavioral errors", yet they both claim to be on the same side. That just shows that these are just after-the-fact explanations as to why a historical pattern is all but guaranteed to repeat in the future and why everybody else needs to jump in and buy the stocks they already own.

--There really hasn't been any real historical research on why value companies have outperformed growth ones. Running a mathematical regression is not studying history. It's not really possible to evaluate the possible truth value of backpacker's original proposal because we really don't have good historical research. That historical research is hard work and doesn't pay well. Starting a mutual fund based on "historical data" is much easier and more profitable. Hence, we have the later and not the former.

One additional comment: More and better data has impacted the way every part of our economy works. It has provided for better decision making in many ways. I don't find it credible to claim that more and better data has not helped investors with valuing companies. This has nothing to do with behavioral errors or risk--I'm just talking about collecting information and determining what price you think something should sell at. Additionally, dumb money (i.e. individual investors) is increasingly indexing as opposed to picking stocks themselves. That means that valuations are being determined by people who do it for a living and not by those who took a trip to the library on Saturday morning to look at Value Line. Those people doing their Saturday research weren't dumb; they simply didn't have access to the information that firms who spend millions on data feeds do.

To argue that the value premium will not decline does seem to argue that any data beyond calculating a simple price-to-X ratio is useless for decision making. Is there really bell curve where using 0 data points is bad, using exactly 2 is ideal, and anything more than 2 pieces of data is harmful?

User avatar
JoMoney
Posts: 7672
Joined: Tue Jul 23, 2013 5:31 am

Re: Better growth? Or on how the value premium ends

Post by JoMoney » Fri Aug 19, 2016 11:26 pm

backpacker wrote:...
I had always thought that shrinking the value premium without shrinking the value/growth spread would require the market to have companies with better growth prospects than in the past. But that's not needed at all. We don't need a market with better companies, we just need a market that's better at sorting the companies it already has into those that will have higher growth and those that will have lower growth. That's enough to shrink the value premium without shrinking the spread.

That the market is getting better at predicting growth isn't a ridiculous idea. It could be that markets have more information than they used to about companies or, more plausibly, markets are better at using the information they already have. The value premium could just be one more casualty of improved market efficiency.
...
It could also be, that assets selling below liquidation value of the assets used to be a 'thing' that no longer exists in the market.
So the market may not be any better at pricing future growth expectations, but maybe in the past mixed in with the low-growth expectation stocks there was once assets priced below what they could be liquidated for outside of the stock market, and those have become a less prevalent factor.

If you were trying to get something extra out of your stock picking (or stock segment/style picking) your options might be something like

(a) Buy assets that you know are worth less than there present liquidation value, and then liquidate (or hope someone else comes along and has the ability to force a liquidation) before the excess value depreciates or is destroyed by poor management

(b) Buy assets that the market has mispriced the growth expectations on relative to what will actually occur

With option (a), clearly the higher the market price multiple the less any under-pricing of the net current assets there is.
Whereas wiith option (b), the relative market price multiple doesn't give you any indication as to how wrong the price is relative to future growth.
The market may be just as horrible at predicting future growth as ever, with both high and low multiple stocks being priced poorly and with equal probability of being wrong with regard to the actual future growth... but what changed, is the disappearance of intrinsically cheap stocks selling at less than net present asset value.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

larryswedroe
Posts: 15989
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Better growth? Or on how the value premium ends

Post by larryswedroe » Sat Aug 20, 2016 7:47 am

stlutz
I have actually said repeatedly that there is a strong risk story, with lots of research showing clear and simple risk based explanations, AND that there is also strong evidence of behavioral explanations and limits to arbitrage which prevents sophisticated investors from fully correcting errors. And thus it's not black or white, it's some of both.

Now risk cannot be arbed away, so if you believe it's risk than you should believe it will continue---though the size of the premium is time varying and can shrink just as the ERP has shrunk. If you believe it's behavioral (or a part of it), then you have to decide if the behaviors are likely to change and/or will limits to arbitrage shrink or disappear.

Larry

User avatar
nedsaid
Posts: 12668
Joined: Fri Nov 23, 2012 12:33 pm

Re: Better growth? Or on how the value premium ends

Post by nedsaid » Sat Aug 20, 2016 1:29 pm

stlutz wrote:A few things which this thread highlights that I'd like to point out:

--From the perspective of stock picking, "risk" vs "behavioral" explanations of historical anomalies are a distinction without a difference. In political economics, the efficient market and behavioral schools have huge disagreements that matter. In this thread, Larry talks about "risk" while nedsaid talks about "behavioral errors", yet they both claim to be on the same side. That just shows that these are just after-the-fact explanations as to why a historical pattern is all but guaranteed to repeat in the future and why everybody else needs to jump in and buy the stocks they already own.

Nedsaid: If you have a better explanation, you are welcome to offer it. Quite frankly, markets do some pretty strange things and it boggles my brain to explain why. Who knows what some large hedge fund manager is doing that could in the future cause a market panic? There is always a big unknown out there that can change everything, at least in the short term.

But after years of watching my own investments under many types of market conditions, the behavioral arguments make sense. The academic research makes sense as it seems consistent with what I know about human behavior and what I have observed about markets.

But as I said above, stuff happens and quite often we really don't know why. The talking heads will mention something about the Fed or a comment some important person made or reaction to an economic number but often they are just searching for an explanation.

The quants crank away 24/7 on their computers and will describe the Value effect in terms of risk and will try to quantify it. In terms of volatility, Value tends to be less volatile than growth. So the risk story doesn't fully satisfy me. There does seem to be a fundamental risk: more volatile earnings, more debt on the balance sheet, uncertain prospects.

But I suppose the quants are not satisfied with the behavioral explanation either.


--There really hasn't been any real historical research on why value companies have outperformed growth ones. Running a mathematical regression is not studying history. It's not really possible to evaluate the possible truth value of backpacker's original proposal because we really don't have good historical research. That historical research is hard work and doesn't pay well. Starting a mutual fund based on "historical data" is much easier and more profitable. Hence, we have the later and not the former.

One additional comment: More and better data has impacted the way every part of our economy works. It has provided for better decision making in many ways. I don't find it credible to claim that more and better data has not helped investors with valuing companies. This has nothing to do with behavioral errors or risk--I'm just talking about collecting information and determining what price you think something should sell at. Additionally, dumb money (i.e. individual investors) is increasingly indexing as opposed to picking stocks themselves. That means that valuations are being determined by people who do it for a living and not by those who took a trip to the library on Saturday morning to look at Value Line. Those people doing their Saturday research weren't dumb; they simply didn't have access to the information that firms who spend millions on data feeds do.

Nedsaid: Actually, I think the more and better data have made the markets more efficient. Also trading volumes have increased which make markets more liquid and probably more efficient. But what you are missing is that markets are not always rational as human emotion is involved. In early 2000, forward market P/E's were 32-35. Fast forward a decade when earnings probably doubled, P/E's were down to maybe 20 or less. I suppose earnings growth prospects were less in 2010 than early 2000, or at least the perceptions of those earnings prospects. But was a dollar of forward earnings in 2010 really worth 50%-60% of what they were valued in 2000?

It is much easier to project out the economic or business return of stocks but who knows what the speculative return (P/E expansion or contraction) will be? Pretty much, speculative return is investor expectations and investor moods. How do you quantify investor moodiness?

This is why I lean on the side of the behavioral argument.


To argue that the value premium will not decline does seem to argue that any data beyond calculating a simple price-to-X ratio is useless for decision making. Is there really bell curve where using 0 data points is bad, using exactly 2 is ideal, and anything more than 2 pieces of data is harmful?


Nedsaid: I don't think there is lack of data, indeed we seem to be swimming in it. There gets to be sort of a Rorschach effect, after a while you see what you want to see. Small Growth is supposed to be the worst asset class ever but the actual Vanguard Index actually has performed pretty well. Then they say, well Vanguard has taken out the lottery stocks. Then somebody could say, well Small Value has done great since 1926 but when could you actually have invested in it? It is sort of like now you see it, now you don't. The academics say that the Small Value premium exists but then Bogle turns around and says that during much of the 1926-2016 period a real life investor could have not invested in it. A real life vs ivory tower argument. Now you see it, now you don't. That is what is so frustrating about this.
A fool and his money are good for business.

larryswedroe
Posts: 15989
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Better growth? Or on how the value premium ends

Post by larryswedroe » Sat Aug 20, 2016 2:06 pm

There really hasn't been any real historical research on why value companies have outperformed growth ones.
this statement is without question false. As I have stated there are more than a dozen papers that have looked at this exact issue and provided logical and simple risk based explanations. And I have listed the papers many times.
larry

User avatar
nedsaid
Posts: 12668
Joined: Fri Nov 23, 2012 12:33 pm

Re: Better growth? Or on how the value premium ends

Post by nedsaid » Sat Aug 20, 2016 4:08 pm

larryswedroe wrote:
There really hasn't been any real historical research on why value companies have outperformed growth ones.
this statement is without question false. As I have stated there are more than a dozen papers that have looked at this exact issue and provided logical and simple risk based explanations. And I have listed the papers many times.
larry
Gosh, wasn't there the Cowles Commission? Wasn't that the precursor to modern academic research? As I recall, they went back as far as they could to compile market stats.
A fool and his money are good for business.

larryswedroe
Posts: 15989
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Better growth? Or on how the value premium ends

Post by larryswedroe » Sat Aug 20, 2016 4:28 pm

nedsaid
As far as I know they just had data, not research attempting to explain the data. Now we have many papers explaining clear and simple risk based explanations for the value premium, which I have posted the list here many times. And my new book, Your Complete Guide to Factor Based Investing will provide a summary of that research for each of the factors we recommend considering
Best wishes
Larry

User avatar
Topic Author
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: Better growth? Or on how the value premium ends

Post by backpacker » Sat Aug 20, 2016 4:51 pm

I thought of a more visual and realistic way to illustrate how a behavioral value premium could go away without shrinking the spread. Imagine that there is a market that looks like this:

Code: Select all

2% 3% 4% | 5% 6% 7% 
   H  H    H  H  H
L  L  L    L  L 
C  C  C    C  C  C
This market has pricing errors. It has five overpriced stocks (the Hs for high) that the market assumes will each grow 1% faster than it should. Their position indicates that the market has priced them to grow between 2-7% respectively. There are also five under priced stocks (the Ls for low), stocks that the market has priced to grow 1% lower than it should. Their position shows that the market has priced them to grow 1-6%. The Cs are stocks that are priced correctly and are there to illustrate that their presence doesn't matter.

The line divides the growth stocks on the right (with predicted growth 5-7%) from the value stocks on the left (with predicted growth 2-4%). All stocks have the same past earnings, so the number of stocks at each predicted growth point fully fixes the average valuation of value stocks and growth stocks. All you have to do is count.

This mispricing creates a value premium. This because value has more underpriced stocks than overpriced stocks and the reverse is true for growth.

Now suppose the market figures out how to eliminate those errors. It shifts the Hs (the overpriced stocks) one step to the left and the Ls (the underpriced stocks) one step to the right.

Code: Select all

2% 3% 4% | 5% 6% 7% 
H  H  H    H  H
   L  L    L  L  L
C  C  C    C  C  C
You can see just by inspection that the overall distribution (i.e. the number of stocks at each predicted growth level) has not changed, so the value spread has not changed. But now there are no pricing errors, so no behavioral value premium. The premium can disappear without the spread shrinking.
larryswedroe wrote: Now back to your example. As I stated the market is not that inefficient.
One advantage of the above way of explaining the idea, as compared to my earlier cases, is that the market doesn't have to be very inefficient. It only needs to be 1% off for the growth of any particular stock. With more stocks and more price points, the errors could be made as small as you like.

larryswedroe
Posts: 15989
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Better growth? Or on how the value premium ends

Post by larryswedroe » Sat Aug 20, 2016 5:15 pm

backpacker, sorry but that is wrong. I'll try again for you.

If the market changes its forecasts the valuations MUST change. The companies with now lower growth will have lower p/es and the companies with higher growth will have higher P/Es and the spread will narrow. Very simple. What do you think happens when companies have disappointing earnings announcements. The prices fall. And the reverse when the beat estimates.

Just ask yourself if you thought a company would grow at 6%, and now you learn it grows at 5%, you will clearly pay less for it. So the P/E falls and the P/B falls. Now you were looking at company expecting it to grow 1% and learn it is growing 2%, so you pay more for it. So it's price rises and it's P/E rises and it's P/B rises and the spread narrows.

Again you don't even need this which is quite simple. all one needs is the cost of capital story.

the way one values a company is to take the earnings expected and that is the numerator, you divide by the risk free rate plus some premium for risk. That determines the NPV, or current price. So doesn't matter what the G is you put in. All the G does is determine the price, not the rate of return which is solely determined by the risk premium, the denominator. That's the company's cost of capital and the expected return. So since growth companies are riskier and require a higher risk premium they must have higher expected returns. All a more accurate forecasting would do is lower the spread.

We know that value companies have more volatile earnings, and lots of other riskier characteristics which are simple and intuitive. If you don't agree suggest you read the literature and explain why it's all wrong. Or simply explain as I have asked before why you think companies with more leverage, more volatility of earnings, more irreversible capital, fewer sources of capital, more at risk of tightening monetary policy and more, are less risky. If you cannot this whole discussion should end because riskier companies should have higher risk premiums required and thus higher expected returns.

This is all basic finance 101.

You will eventually see that you're incorrect.
Larry

lemonPepper
Posts: 89
Joined: Wed Feb 18, 2015 7:43 pm

Re: Better growth? Or on how the value premium ends

Post by lemonPepper » Sat Aug 20, 2016 5:39 pm

Larry, I definitely appreciate you trying to explain this in many different ways.

User avatar
Topic Author
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: Better growth? Or on how the value premium ends

Post by backpacker » Sat Aug 20, 2016 5:45 pm

larryswedroe wrote:The companies with now lower growth will have lower p/es and the companies with higher growth will have higher P/Es...
Yes.
larryswedroe wrote: ...and the spread will narrow.
No.

Just look at the diagrams. At every valuation level (there represented as growth estimates), an overpriced stock that moves down is replaced by underpriced stock that moves up. The PE ratios of each of those stocks individually changes, sure, but the movements offset at the aggregate level. Nothing happen to the average valuations. Just count the number of stocks at each valuation level (i.e. the number of stocks in each column). It stays the same.
larryswedroe wrote: Again you don't even need this which is quite simple. all one needs is the cost of capital story.
You can add variable cost of capital to my story if you like. Maybe the market uses a 5% discount rate for growth stocks (i.e. anything on the right side of the line) and a 7% discount rate for value stocks (anything on the left). The value premium generated by different discount rates will survive of course, but the mispricing layered on top of it (i.e. the free stop at the desert tray) will go away without narrowing the spread.
stlutz wrote: There really hasn't been any real historical research on why value companies have outperformed growth ones. Running a mathematical regression is not studying history. It's not really possible to evaluate the possible truth value of backpacker's original proposal because we really don't have good historical research.
This is a nice observation and one that I agree with. While there may be plenty of people digging through history to generate hypotheses about the value premium, there's no consensus on the precise mechanism. What this exercise has taught me, if anything, is that the precise mechanism matters for things like whether or not we can use value spreads to figure out if the value premium will be higher or lower in the past or even to figure out if its still there at all. Even if you say that the premium is generated by pricing errors, well, what sort? How do they get in? How are they distributed? This stuff matters.

larryswedroe
Posts: 15989
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Better growth? Or on how the value premium ends

Post by larryswedroe » Sat Aug 20, 2016 6:02 pm

packer, forget your diagram, doesn't show anything. You just agreed the spread narrows. That should end the discussion
And as to the risks none of them are diversifiable, so unless you believe it's all behavioral and it will all go away the value premium will stay and the spreads narrow under either scenario

And clearly you don't understand the cost of capital story, it's simple math as I stated. The value premium must persist if the discount rate is larger.

Yes there is no consensus as I said, but there are clear arguments on both sides, showing it's both, risk and behavioral with some of both.
And it must be that the value spreads matter, why do you think the leading finance people talk about it. Do you believe you know something they don't?

sorry but this really is my last attempt. Just don't have time to explain further
But I'm confident if you actually have open mind you will eventually agree you must be wrong.


Larry

User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Better growth? Or on how the value premium ends

Post by in_reality » Sat Aug 20, 2016 6:10 pm

Whether it's behavior or risk, I know I was buying more shares of energy stocks as they tanked recently.

It's a value strategy that indexes weren't following, so even if individual investors aren't buying hot growth stocks on their own and using indexes instead, they'd still be putting less money into energy with each new purchase as energy stocks dropped and more into energy with each new purchase as energy stocks rose.

This is both additionally risky on my part and not what investors typically do.

I don't pretend to know if any benefit from such a strategy will overcome costs but it suits me as an investor.

User avatar
Topic Author
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: Better growth? Or on how the value premium ends

Post by backpacker » Sat Aug 20, 2016 6:22 pm

larryswedroe wrote:You just agreed the spread narrows.
Where? I said no such thing.
larryswedroe wrote: This really is my last attempt. Just don't have time to explain further...
I'm not convinced you grasp the issues well enough to have anything to explain. Your job is to sell the value premium using information other people have given you, not study the value premium. I was too optimistic when I thought you would think carefully about my ideas. :?
Last edited by backpacker on Sat Aug 20, 2016 6:27 pm, edited 1 time in total.

User avatar
JoMoney
Posts: 7672
Joined: Tue Jul 23, 2013 5:31 am

Re: Better growth? Or on how the value premium ends

Post by JoMoney » Sat Aug 20, 2016 6:26 pm

in_reality wrote:Whether it's behavior or risk, I know I was buying more shares of energy stocks as they tanked recently.

It's a value strategy that indexes weren't following, so even if individual investors aren't buying hot growth stocks on their own and using indexes instead, they'd still be putting less money into energy with each new purchase as energy stocks dropped and more into energy with each new purchase as energy stocks rose.

This is both additionally risky on my part and not what investors typically do.

I don't pretend to know if any benefit from such a strategy will overcome costs but it suits me as an investor.
This is basicly a mean-reversion strategy vs a momentum strategy.
Regardless of any costs in trading, what will make the difference is future performance. If what's been doing well continues to do well (or if what's been doing poorly continues to do poorly), your relative results will suffer. If the relative directions change and the relative divergence starts to converge it will work out slightly better.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

User avatar
Topic Author
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: Better growth? Or on how the value premium ends

Post by backpacker » Sat Aug 20, 2016 6:36 pm

For those interested, my way of thinking about the behavioral value premium in this thread is basically just Arnott and Hsu's. They think the value premium is generated by adding "noise" to fair prices.

http://papers.ssrn.com/sol3/papers.cfm? ... _id=928167

larryswedroe
Posts: 15989
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Better growth? Or on how the value premium ends

Post by larryswedroe » Sat Aug 20, 2016 8:28 pm

Backpacker
You agreed with this statement below which means you must agree that the spread has narrowed. If you don't get that it's hopeless I'm afraid. But I'll give simple example, say all value stocks are trading at PE of 10 and all growth at 20 and now you correct the growth estimates and lower for growth and higher for value. So now growth stocks trade at 19 and value at 11.And the spread narrows.
The companies with now lower growth will have lower p/es and the companies with higher growth will have higher P/Es...
And note even if you buy the behavioral story, which I already said it is PARTLY the rational, the spread must narrow or the premium will not disappear. Again, cost of capital. Nothing more you need.

Also Arnot and Hsu ignore all the simple and common ECONOMIC rational. Again if you disagree that the characteristics I mentioned, which are well documented as characteristics that are common to value companies are not risky explain why you think that, and non of those characteristics are diversifiable, repeating it. Arnott is trying simply to sell his indices, which are nothing more as Asness showed clearly a value tilt.
Larry

User avatar
FIREchief
Posts: 3576
Joined: Fri Aug 19, 2016 6:40 pm

Re: Better growth? Or on how the value premium ends

Post by FIREchief » Sat Aug 20, 2016 8:42 pm

larryswedroe wrote: But I'm confident if you actually have open mind you will eventually agree you must be wrong.
Signature material?? :thumbsup
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.

User avatar
Topic Author
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: Better growth? Or on how the value premium ends

Post by backpacker » Sat Aug 20, 2016 10:26 pm

larryswedroe wrote: You agreed with this statement...
The companies with now lower growth will have lower p/es and the companies with higher growth will have higher P/Es...
Yes.
larryswedroe wrote: ...which means you must agree that the spread has narrowed
No.

There are exactly two stocks A and B with the same trailing earnings. At the close on Monday, stock A has a PE of 21 and B a PE of 20. Since they are the only stocks, growth (top half of the market by PE) has an average PE of 21. Value (the bottom half of the market by PE) has an average PE of 20. Overnight, the market changes its mind about the growth prospects of the two companies so, at the start of trading the next day, the stocks are switched. B has a PE of 21 and A has a PE of 20.

Obviously, stock A has a lower PE on Tuesday and B a higher PE. But the average PE of the top half of the market is 21 throughout and the average PE of bottom half of the market is 20 throughout.

This is not difficult.

For a less contrived and more realistic example, see my post above with charts.
larryswedroe wrote: But I'll give simple example, say all value stocks are trading at PE of 10 and all growth at 20 and now you correct the growth estimates and lower for growth and higher for value. So now growth stocks trade at 19 and value at 11. And the spread narrows.
I agree that if you uniformly shift the PE of growth stocks down and uniformly shift the PE of value stocks up (and don't shift the PE of growth stocks below that of value stocks), then yes. The spread narrows. But valuations need not shift uniformly and my case with the charts, if you take the time to understand it, is explicitly one in which the PE of growth stocks is not moving uniformly or even all in the same direction.
larryswedroe wrote: And note even if you buy the behavioral story, [...] the spread must narrow or the premium will not disappear. Again, cost of capital. Nothing more you need.
Saying that cost of capital explains the return of value stocks is like saying that earnings estimates explain the price of value stocks. Not false, but not useful. The cost of capital explains the price of literally everything. It's just the discount rate investors use to set prices after estimating earnings, which could in theory be set by their awareness of risk factors or their behavioral bias or by a chimp throwing darts. Cost of capital is just a term for number that gets plugged into a discount model.

What I think you mean is that a fully rational risk-based cost of capital story can do all the work. Sure, but then where is your supposed free stop at the desert tray? To get that, you need to get pricing errors in there somehow.

I'm showing that the market can eliminate pricing errors, and with them your free lunch, without narrowing the value spread. Who cares about the risk part. That's a different issue.

User avatar
PeteyDink
Posts: 86
Joined: Fri May 06, 2016 5:18 pm

Re: Better growth? Or on how the value premium ends

Post by PeteyDink » Sun Aug 21, 2016 1:18 am

If growth stocks have a high price to earnings ratio, and a high P/E indicates positive investor sentiment, is value investing a contrarian strategy?

User avatar
JoMoney
Posts: 7672
Joined: Tue Jul 23, 2013 5:31 am

Re: Better growth? Or on how the value premium ends

Post by JoMoney » Sun Aug 21, 2016 2:37 am

PeteyDink wrote:If growth stocks have a high price to earnings ratio, and a high P/E indicates positive investor sentiment, is value investing a contrarian strategy?
The fact that there are a multitude of mutual funds catering to people who want to follow such strategies is an indicator that it's not a contrarian strategy. Contrarian investing isn't just following an opposing strategy, it's being different than the crowd. I don't think it's even fair to call what people do around here "value investing". Buying a low P/E stock doesn't make it a value, there could be a higher P/E stock being offered at a much better price relative to what it's worth. The question should be whether you're able to determine what a stock is worth better than the market.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

User avatar
FIREchief
Posts: 3576
Joined: Fri Aug 19, 2016 6:40 pm

Re: Better growth? Or on how the value premium ends

Post by FIREchief » Sun Aug 21, 2016 2:49 am

JoMoney wrote: Buying a low P/E stock doesn't make it a value, there could be a higher P/E stock being offered at a much better price relative to what it's worth. The question should be whether you're able to determine what a stock is worth better than the market.
Bingo! And, of course we can't determine any such thing..... :beer
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.

larryswedroe
Posts: 15989
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Better growth? Or on how the value premium ends

Post by larryswedroe » Sun Aug 21, 2016 8:08 am

backpacker
Again don't know why you cannot see how your example is wrong. In your example the expected return of A at the start say is 4.8 and the return of B is 5. Now the spread predicts the expected return difference. Now they reverse, A now has expected return of 5 and B it's 4.8. The spread is the same and the expected return is the same. The value premium is the same. Quite simple as I said.
And it's clear you don't understand the cost of capital issue, so I'm not going to go into any longer, just leaving you with this data that shows it. The cost of capital explains the return of all stocks. And we have the evidence, returns are monotonic going from deciles whether we look at P/E or CAPE 10 or BTM or P/CF or any other metric like that.


I would add that DFA has a paper on the value premium and specifically identifies that the spread is the measure of the expected return gap, only it's not useful for timing. But the wider the spread the narrower the premium and vice versa. Now they have a Nobel Prize winner who ran the research time and bunch of world class finance people. AQR also uses the spread in say way, and they have a bunch of the leading finance professors, but you think you've got this right and are they are all wrong? Rhetorical

I'll just end with this, it ain't what a man don't know that gets him in trouble, it's what he knows for sure but ain't so.
Like I said eventually you'll figure it out that your just wrong. I'm done
Larry

User avatar
nedsaid
Posts: 12668
Joined: Fri Nov 23, 2012 12:33 pm

Re: Better growth? Or on how the value premium ends

Post by nedsaid » Sun Aug 21, 2016 10:08 am

larryswedroe wrote:backpacker, sorry but that is wrong. I'll try again for you.

If the market changes its forecasts the valuations MUST change. The companies with now lower growth will have lower p/es and the companies with higher growth will have higher P/Es and the spread will narrow. Very simple. What do you think happens when companies have disappointing earnings announcements. The prices fall. And the reverse when the beat estimates.

Nedsaid: Larry, I agree with you on this one. All other things being equal, higher forecasted earnings cause prices to rise and lower forecasted earnings call prices to fall. Market forecasts of future earning are a primary driver of stock prices.

There are other factors involved such as the level of interest rates, lower interest rates tend towards higher P/E's and higher interest rates tend towards lower P/E's.

Another factor is investor optimism or pessimism or better defined as speculative return. Market optimism tends toward higher P/E's and market pessimism tends towards lower P/E's.


Just ask yourself if you thought a company would grow at 6%, and now you learn it grows at 5%, you will clearly pay less for it. So the P/E falls and the P/B falls. Now you were looking at company expecting it to grow 1% and learn it is growing 2%, so you pay more for it. So it's price rises and it's P/E rises and it's P/B rises and the spread narrows.

Nedsaid: It is a little more complicated than that. There is the aspect of financial engineering where the management manages earnings expectations and the stock price more than the actual businesses.

For example, TYCO showed an artificially high earnings growth because of all the acquisitions. WorldComm was a similar story. If you took the increase of earnings of the old and newly acquired businesses in aggregate, earnings growth would have been a lot lower. It was a comparison of the earnings from the old business versus the earnings of the old and new businesses. Whatever earnings increases from the combined businesses were mainly from firing people and not from synergies. In many instances, the synergies never materialized because of the clash of corporate cultures and the loss of many of the best people in the acquired companies.

GE would time asset sales to help make the steady growth of earnings more than what it really was, the company also transformed itself from a boring, old industrial company into a bank. GE also kept spinning off or selling its lower margin businesses as did Coke. I used to joke that Coke would eventually spin everything off but the logo and then make the logo 15% bigger every year to maintain earnings growth!

So a lot of "earnings growth" were growth of estimates based on a CEO's glowing statements and not so much organic growth of the underlying businesses. The late Coke CEO used to phone and berate analysts who questioned Coke's alleged 15% growth rate.

Again you don't even need this which is quite simple. all one needs is the cost of capital story.

the way one values a company is to take the earnings expected and that is the numerator, you divide by the risk free rate plus some premium for risk. That determines the NPV, or current price. So doesn't matter what the G is you put in. All the G does is determine the price, not the rate of return which is solely determined by the risk premium, the denominator. That's the company's cost of capital and the expected return. So since growth companies are riskier and require a higher risk premium they must have higher expected returns. All a more accurate forecasting would do is lower the spread.

Nedsaid: Larry, you have me until the statement above. You said, "So since growth companies are riskier and require a higher risk premium they must have higher expected returns." Then below you say, "We know that value companies have more volatile earnings, and lots of other riskier characteristics which are simple and intuitive." There is a paradox here. You must be saying that the higher expected returns of Growth stocks tend not to materialize. There cannot be both higher expected returns from Growth stocks and a Value premium unless the earnings forecasts for Growth stocks are always too high. Conversely, the earnings forecasts for Value stocks must always be too low.

So this really plays into the behavioral argument that I make. Investors are too optimistic about Growth stocks and too pessimistic about Value stocks. Probably a big reason for the over optimism for Growth is the financial engineering that I pointed out above.


We know that value companies have more volatile earnings, and lots of other riskier characteristics which are simple and intuitive. If you don't agree suggest you read the literature and explain why it's all wrong. Or simply explain as I have asked before why you think companies with more leverage, more volatility of earnings, more irreversible capital, fewer sources of capital, more at risk of tightening monetary policy and more, are less risky. If you cannot this whole discussion should end because riskier companies should have higher risk premiums required and thus higher expected returns.

Nedsaid: Value companies tend to be less volatile over time than Growth companies. That is what my eyeballing has told me. But you make a good argument for greater fundamental risk of Value companies. But again, you need to explain the paradox that I raised above. Pretty much, Growth companies have higher expected returns because of risk and there is a Value premium because of risk. Perhaps it could be said that the additional return from the additional fundamental risks of Value stocks outweighs the higher future expected returns of Growth stocks from additional market risk.

As Desi Arnaz said to Lucy, "You've got some 'splaining to do."


This is all basic finance 101.

You will eventually see that you're incorrect.
Larry
A fool and his money are good for business.

larryswedroe
Posts: 15989
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Better growth? Or on how the value premium ends

Post by larryswedroe » Sun Aug 21, 2016 12:24 pm

nedsaid
Yes I agree that it can be more complicated, so in this type discussion you assume all else equal without having to repeat it all the time.
With that said, that is why we have some anomalies that get discovered, so there is the well known accrual anomaly, and that is also why people like cash flow or EBITDA measures relative to price and why quality matters too. So that is how you end up with behavioral anomalies until they get discovered, and then limits to arb sometimes allow them to continue, at least in small stocks.

Re volatility, value companies are actually more volatile in the long term. People who like to make the case that it's all behavioral like to show only recent data, but if you look at the long term data you find value with higher SD. In LG it's SD of about 20 and LV closer to 27. In small both are about 32. But we know about the sg black hole. And with large dominating the market cap overall value less volatile. Now we also know that volatility isn't only form of risk. Just one measure. Other things count too like higher moments and when the risks tend to show up. so example, value companies have more irreversible capital so their risks tend to show up in bad times, and investors demand premiums for taking risks that show up in bad times, which is why stocks have such a high premium.

Sorry on your growth anomaly, I don't know what you are referring to, growth stocks do not have higher expected returns, not in any model I know.

Larry

User avatar
PeteyDink
Posts: 86
Joined: Fri May 06, 2016 5:18 pm

Re: Better growth? Or on how the value premium ends

Post by PeteyDink » Sun Aug 21, 2016 12:42 pm

hominy wrote: The fact that there are a multitude of mutual funds catering to people who want to follow such strategies is an indicator that it's not a contrarian strategy. Contrarian investing isn't just following an opposing strategy, it's being different than the crowd. I don't think it's even fair to call what people do around here "value investing". Buying a low P/E stock doesn't make it a value, there could be a higher P/E stock being offered at a much better price relative to what it's worth. The question should be whether you're able to determine what a stock is worth better than the market.

My definition of value investing was simplified for the question, it's not just about P/E of course.

However, just because value funds exist doesn't mean they aren't contrarian. There is more money outside of value equity than in it. Arguments about factor loadings exist because people disagree on their utility and need. Part of the reason why it's called value is that it's part of the market that isn't as popular, isn't sexy from a growth standpoint. Most of the market follows emotion and greed. I'm not positive, but I also expect that most institutional investors, the ones who impact market changes the most, usually do not have a strategy with a large allocation to value funds. (Please correct me if I'm wrong).

Bogleheads and others talk openly about value but most people don't know or don't care.

User avatar
nedsaid
Posts: 12668
Joined: Fri Nov 23, 2012 12:33 pm

Re: Better growth? Or on how the value premium ends

Post by nedsaid » Sun Aug 21, 2016 3:33 pm

larryswedroe wrote:nedsaid
Yes I agree that it can be more complicated, so in this type discussion you assume all else equal without having to repeat it all the time.
With that said, that is why we have some anomalies that get discovered, so there is the well known accrual anomaly, and that is also why people like cash flow or EBITDA measures relative to price and why quality matters too. So that is how you end up with behavioral anomalies until they get discovered, and then limits to arb sometimes allow them to continue, at least in small stocks.

Nedsaid: Good point about cash flow, it eliminates the issue of accruals.

I have been skeptical of EBITDA (Earnings before interest, taxes, depreciation, amortization). I jokingly refer to it as Earnings before Expenses. Interest and taxes are true cash expenses, I have often felt that EBITDA was a crutch used by companies like Cable companies that have high investment and often low earnings. If you have to use EDITDA, you are admitting you don't have a very good business or at least one that is very capital intensive. It makes as much sense to me as saying earnings before employee compensation.

Cash flow does take into account non-cash charges like depreciation and amortization and in effect adds them back to earnings and that makes sense. Depreciation is a real expense as it represents the wear and tear on assets that will likely have to be replaced sometime in the future, but it doesn't cost you any cash right now. Depreciation is like saying that you are setting aside cash for future replacement. Cash flow is a good metric as long as one knows how it is arrived at.


Re volatility, value companies are actually more volatile in the long term. People who like to make the case that it's all behavioral like to show only recent data, but if you look at the long term data you find value with higher SD. In LG it's SD of about 20 and LV closer to 27. In small both are about 32. But we know about the sg black hole. And with large dominating the market cap overall value less volatile. Now we also know that volatility isn't only form of risk. Just one measure. Other things count too like higher moments and when the risks tend to show up. so example, value companies have more irreversible capital so their risks tend to show up in bad times, and investors demand premiums for taking risks that show up in bad times, which is why stocks have such a high premium.

Nedsaid: I must have seen data from recent years. How far back do you have to go and how long of time periods do you need to see the extra volatility from Value? Not disputing what you are saying but I haven't seen this myself. It would make sense that if Value stocks have more fundamental risk that they would also be more volatile. Thanks for the information.

Sorry on your growth anomaly, I don't know what you are referring to, growth stocks do not have higher expected returns, not in any model I know.

Nedsaid: Larry, as far as the growth anomaly, you said it yourself. See below.
So since growth companies are riskier and require a higher risk premium they must have higher expected returns.
Then later on you say this:
We know that value companies have more volatile earnings, and lots of other riskier characteristics which are simple and intuitive. If you don't agree suggest you read the literature and explain why it's all wrong. Or simply explain as I have asked before why you think companies with more leverage, more volatility of earnings, more irreversible capital, fewer sources of capital, more at risk of tightening monetary policy and more, are less risky. If you cannot this whole discussion should end because riskier companies should have higher risk premiums required and thus higher expected returns.



The wires got crossed a bit there in your post and it needs further explanation. Please re-read your own post. I thought I made this crystal clear. I pointed out a paradox in your statements but reconciled it by wondering if both earnings forecasts and future expected returns of growth stocks consistently fall short. I then theorized that earnings forecasts for Value stocks as a whole are consistently too low. If you can't see what I am getting at, please read your post and my response again.

You say that growth stocks are riskier and investors demand higher return. Then you say the same thing about value stocks. You used the phrase "higher expected returns" for both. I cannot be the only person in the world who is seeing the paradox in your statements.

Doesn't mean what you said was wrong but that it needs further clarification.

Thanks.



Larry
A fool and his money are good for business.

User avatar
nedsaid
Posts: 12668
Joined: Fri Nov 23, 2012 12:33 pm

Re: Better growth? Or on how the value premium ends

Post by nedsaid » Sun Aug 21, 2016 4:38 pm

Another concept that has helped me is the concept of a growth stock.

To me, a growth stock is a company that is growing faster than the economy. Let's say that the economy now grows at 2% and inflation is 1%. By my definition, a growth stock would have to grow earnings by more than 3% a year. In more normal times, the economy would grow at 3% and inflation would be 2%. Probably 5-6% would be considered slow growth. 8% - 10% a year consistent earnings growth gets you into blue chip territory. 15-20% a year consistent growth gets you into the fast growth territory.

Perceptions change slower than reality. Coke and GE wanted to be in the fast growth category when in reality, they were blue chips. The CEOs of both companies worked very hard to manage earnings because they realized that the drop in perception from fast grower to blue chip would mean a drop in P/E multiples, thus a drop in stock price, and a drop in CEO compensation through stock options and grants.

Another couple concepts that help are that the more consistent and sustainable the earnings growth is, the greater a premium that the market puts upon the stock. Wall Street loves consistent growth.

Faster growth rates will tend towards higher P/E's. Nirvana is when you get consistent 15%-20% earnings growth that is sustainable. You could get P/E's for these stocks in the thirties and forties. This is why General Electric had a 45 P/E at one time. Growth was consistent and perceived to be 15% a year. My guess is that earnings growth minus all the financial engineering was probably about 8%, still pretty good for a very large company. GE should have probably traded at the very most a P/E of 25.

So that is the trick of earnings estimates. As a CEO, you want the estimates to be higher and not lower. But if the company isn't really performing up to the estimates, it takes a while for analysts to realize that delivering upon the higher estimates just isn't in the cards. This is why you see all the games played with earnings numbers. You get EBITDA, operating earnings, pro-forma earnings but anything but GAAP (Generally Accepted Accounting Principles) earnings. It gets to be an important but very silly game. When you have to resort to these gimmicks, the news isn't all good.

You also have to understand that the Brokerage firms have a conflict of interest. If for example, a XYZ brokerage analyst had the gall to say that GE or Coke didn't actually have 15% growth rates, XYZ brokerage management if not the analyst themselves would get an angry call from the CEO. If Coke or GE in the future wanted to float more stock or bonds, good luck with XYZ Brokerage and Investment Bank getting a piece of that business. So the analysts have an incentive to be optimistic. If the Coke CEO says that earnings are growing at 15% and not 8%, the earnings are growing at 15% because he said so and likely your job depends on it.

So all of this needs to be factored in when we discuss earnings expectations, valuations, and future expected returns.

The academic definition would be that Value would be the bottom 30% of stocks by certain valuation measurements: price/earnings, price/cash flow, price/book. Growth stocks would be the top 30%. Core stocks are the ones in the middle. My definitions of Growth and Value might deviate from the academic definitions somewhat.
A fool and his money are good for business.

User avatar
FIREchief
Posts: 3576
Joined: Fri Aug 19, 2016 6:40 pm

Re: Better growth? Or on how the value premium ends

Post by FIREchief » Sun Aug 21, 2016 5:01 pm

Thanks for that great post nedsaid. You've explained things in a way that makes much better sense than many. :sharebeer
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.

User avatar
nedsaid
Posts: 12668
Joined: Fri Nov 23, 2012 12:33 pm

Re: Better growth? Or on how the value premium ends

Post by nedsaid » Sun Aug 21, 2016 10:06 pm

backpacker wrote: Just look at the diagrams. At every valuation level (there represented as growth estimates), an overpriced stock that moves down is replaced by underpriced stock that moves up. The PE ratios of each of those stocks individually changes, sure, but the movements offset at the aggregate level. Nothing happen to the average valuations. Just count the number of stocks at each valuation level (i.e. the number of stocks in each column). It stays the same.
Backpacker, what you are thinking of is that percentages of stocks in the value, core, and growth categories will remain the same. If a stock dropped from Growth to Core, I suppose another stock would move from Core to Growth. Morningstar does that dividing stocks 1/3 value, 1/3 core, 1/3 growth. The academics define Value as the bottom 30% of certain valuation measurements.

The issue that you forget is that the P/E ratios and other financial ratios for the market as a whole can change over time. Most of the time, market P/E's for the US market fluctuate from 10 to 20, 16 is about average. The changing P/E ratios help measure investor enthusiasm. During the 2008-2009 bear market, stocks got hit so hard that some Growth stocks started hitting Value managers stock screens. So both Value and Growth got more valuey. But yet, the proportion of Value stocks, Core Stocks, and Growth stocks remained the same by Morningstar and academic definitions. So valuation measurements can drop not only at the individual stock level but also in aggregate.

Also, if the earnings estimates of one company falls and then its P/E ratio falls it doesn't follow that the earnings estimates of another company rises and its P/E ratio rises. Earnings estimates can drop for big portions of the market partially offset by many fewer companies with increasing earnings estimates. This would happen as the economy goes into recession. If investors get discouraged, P/E ratios could fall even more than what is warranted by decreased earnings estimates.

I have never given too much thought about cost of capital, I always assumed that for stocks it was about 10 percent as that is the historical returns of the stock market and that for bonds it was whatever the market rate is at the time. In other words, I have thought of cost of capital for stocks as the rate of return that investors were expecting or demanding from the stock market.

Larry is thinking of it more as being influenced by interest rates. It is a lot like doing a present value calculation discounting future earnings back to the present using current interest rates. In the formula future earnings/risk free rate + risk premium=intrinsic value. What Larry says makes sense because lower interest rates would cause a lower denominator and thus higher prices. This is what would cause higher P/E's.

So to illustrate, stock A has estimated earnings of $10/share for 2017. Lets put the risk-free rate at 1.5% and the equity risk premium at 4%. That is $10/.055=$181.82 per share. At that price, you would have a P/E ratio of 18. But that would be the intrinsic value of the stock. The actual price would be different.

Just for fun, lets see what stock A would look like in more normal times, lets say a risk-free rate of 4% and an equity premium of 7%. We will use the $10/share earnings estimate for 2017. $10/.11=$90.91. That would give you a P/E ratio of 9.

Am I doing this correctly? You can see that falling interest rates and a lower equity premium drives P/E ratios up because the denominator is decreasing. But with higher P/E's come lower future expected returns.

Hmm. Interesting.
A fool and his money are good for business.

User avatar
Topic Author
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: Better growth? Or on how the value premium ends

Post by backpacker » Sun Aug 21, 2016 11:37 pm

larryswedroe wrote: Again don't know why you cannot see how your example is wrong. In your example the expected return of A at the start say is 4.8 and the return of B is 5. Now the spread predicts the expected return difference. Now they reverse, A now has expected return of 5 and B it's 4.8. The spread is the same and the expected return is the same.
I'm glad we agree on this now. Growth stocks can go down in value and a value stock up in value without changing the spread between growth and value stocks. This because the movements can switch which stocks are growth stocks and which stocks are value stocks.

You want to know why this means that the value premium can shrink. A starts priced for expected return of 5 but, in fact, the market has made a mistake. Given its information, A should be priced for 4.8. That's its fair value. B starts priced for 4.8, but fair value is 5. As long as that persists, there will be an expected .2 value premium created by behavioral errors. A free lunch.

Now the market figures out that it has made a mistake, so reprice A for expected returns of 4.8 and B for expected returns of 5. Since market prices are now fair value prices, there's no behavioral premium anymore. But as you note, the spread hasn't changed.
larryswedroe wrote: And it's clear you don't understand the cost of capital issue, so I'm not going to go into any longer...
You wave around the term "cost of capital" as though it were a magic talisman meant to ward of evil arguments you're not understanding. Stop that. It's not intellectually honest. Ask me questions if you're confused or can't see how the cases work. Or explain what it is you have in mind. I prefer learning things to winning arguments on the internet with people I don't know.
larryswedroe wrote: I would add that DFA has a paper on the value premium and specifically identifies that the spread is the measure of the expected return gap, only it's not useful for timing. But the wider the spread the narrower the premium and vice versa. Now they have a Nobel Prize winner who ran the research time and bunch of world class finance people. AQR also uses the spread in say way, and they have a bunch of the leading finance professors, but you think you've got this right and are they are all wrong? Rhetorical
Because everyone remembers how well things went last time we just assumed "world class finance people" knew what they were doing. :oops:

Look. I would love to be shown wrong by an argument that displayed real understanding of the issues rather than a lame appeal to authority. Do you have one?
Last edited by backpacker on Mon Aug 22, 2016 12:00 am, edited 1 time in total.

User avatar
Topic Author
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: Better growth? Or on how the value premium ends

Post by backpacker » Sun Aug 21, 2016 11:55 pm

Just for fun, note that the value spread can even increase as a behavioral value premium disappears. Suppose there are three stocks whose PEs are as follows (listed under price). Their fair PE ratios given what the market knows are listed under fair.

Code: Select all

  Price Fair
A  20    20
B  19    19
C  18    22
A is the growth stock, B is the neutral stock, and C is the value stock. There is a value premium because investors are getting C for a PE of 18 when it should be at 22. The value spread is 20-18=2.

Now the market fixes the error.

Code: Select all

  Price Fair
C  22    22
A  20    20
B  19    19


The value premium goes away, because everything is fairly valued, but now B is the value stock and C the the growth stock, with A the neutral stock. Now the value spread is 22-19=3.

So the value spread can even increase as any behavioral value premium is eliminated.

User avatar
nedsaid
Posts: 12668
Joined: Fri Nov 23, 2012 12:33 pm

Re: Better growth? Or on how the value premium ends

Post by nedsaid » Mon Aug 22, 2016 1:15 am

Okay, let's look at actual mutual funds and their 15 year performance and standard deviation measurements. Is Value less volatile than Growth? Is there a Value premium? This is just for fun as I am going back 15 years using Morningstar data. In the ratings & risk tab for each fund, their data goes back 15 years.

Plain Old Boring Large Value 7.16% Annual Return 13.55 Standard Deviation
S&P 500 6.08% Annual Return 14.71 Standard Deviation
Income & Growth 5.66% Annual Return 14.77 Standard Deviation
Vanguard Value Index 5.95% Annual Return 15.26 Standard Deviation
Loomis Sayles Value Y 6.90% Annual Return 15.19 Standard Deviation
Dimensional Large Cap Value 7.38% Annual Return 17.95 Standard Deviation

Looks like Larry is right here. These are all funds that I own except for Vanguard Value Index and DFA. All except for my Plain Old Boring Large Value fund had a higher volatility measure than the S&P 500. Only two of the actively managed Value funds beat the S&P 500 Index. Income & Growth is considered a Value fund but the focus is on dividend income higher than the S&P 500, less risk, and value is a secondary consideration. Income & Growth is a quant fund or what I call an optimized index.

Allianz NFJ Small-Cap Value 10.18% Annual Return 15.06 Standard Deviation
Vanguard Small-Cap Value Index 8.79% Annual Return 18.94 Standard Deviation
Vanguard Small-Cap Index 8.81% Annual Return 19.11 Standard Deviation
DFA US Small Cap Value 9.54% Annual Return 21.05 Standard Deviation


Allianz NFJ did a great job beating the Small Index and Small Value Indexes with higher return and
lower Standard Deviation. Vanguard Small-Cap Index and Vanguard Small-Cap Value Index were virtually identical over the 15 year period.

From just eyeballing, I would say Vanguard doesn't do a good job factor loading its indexes for Value. In fact, I would say they do a poor job at it. I would also say that the managers of my Plain Old Boring Large Value fund and Allianz NFJ Small-Cap Value have done an outstanding job.
Last edited by nedsaid on Mon Aug 22, 2016 9:42 am, edited 1 time in total.
A fool and his money are good for business.

larryswedroe
Posts: 15989
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Better growth? Or on how the value premium ends

Post by larryswedroe » Mon Aug 22, 2016 7:34 am

Backpacker,
I'm done, nothing more I can say except your arguments are incorrect. I've tried my best to explain why. so will not be responding any longer. But it's interesting to think that you understand the math and the financial issues better than the leading financial economists. Do you see any single paper or writings that support your view? Does it not even give you pause that you just might be wrong. All rhetorical questions. No need to answer.

And if you are referring to LTCM that has nothing to do with world class finance people ---they IGNORED their own advice and made bets based on opinions and used high leverage. And what's interesting is that all their bets would have played out fine if they had not used leverage, or small leverage, and could have held to maturity (but they were bets that had nothing to do with financial theory).

Larry

bjr89
Posts: 149
Joined: Sat Apr 23, 2016 8:15 am

Re: Better growth? Or on how the value premium ends

Post by bjr89 » Tue Aug 23, 2016 11:47 am

-----
Last edited by bjr89 on Fri Nov 16, 2018 9:00 pm, edited 1 time in total.

User avatar
Topic Author
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: Better growth? Or on how the value premium ends

Post by backpacker » Tue Aug 23, 2016 2:53 pm

nedsaid wrote: Looks like Larry is right here. These are all funds that I own except for Vanguard Value Index and DFA. All except for my Plain Old Boring Large Value fund had a higher volatility measure than the S&P 500.
An interesting observation here nedsaid. One question that I don't know the answer to is how to distinguish different kinds of risk. Whenever you carve off a chunk of the market (like value), you should expect that chunk to have more volatility than the market as a whole. The volatility of an energy sector fund will be higher than the volatility of the S&P for instance, but that doesn't show that energy companies have more risk than average. The energy fund has more volatility only because it is less diversified, not because energy companies are inherently riskier. The question is how much value funds are like that. Are they riskier just because they are less diversified than the total market?

User avatar
Topic Author
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: Better growth? Or on how the value premium ends

Post by backpacker » Tue Aug 23, 2016 3:01 pm

bjr89 wrote:Backpacker, I hear what you're saying, but I think that if the behavioral error portion of the value premium were to go away, it would necessarily result in value spread compression assuming there is a proper, fixed constant for the value spread. Imagine there are and always have been only two stocks in the market. One trades at 10x P/E and the other 40x P/E. If the 10x P/E stock has historically produced return too high relative to its risk, in an efficient market it should be assigned a higher P/E. So maybe 15x P/E is more rational. The opposite for the growth stock. If it has historically produced too low of a return relative to its risk, in an efficient market it should be assigned a lower P/E. So maybe 35x P/E is more rational. Assuming this is the proper price for risk, the behavioral error component has now been eliminated and value spread compression was the result. Adding more stocks to the example doesn't change a thing.
Thanks for a nice post bjr89. I agree that if correcting pricing errors doesn't cause any migration--doesn't turn any value stocks into growth stocks or visa versa--then the spread will compress. That's true in your example. Things can go differently though if correcting pricing errors turns what used to be amazing value stocks into average growth stocks or crappy growth stocks into average value stocks.

Another way to put it is that the spread between value and growth is very different than the spread between stocks and bonds. Stocks can't turn into bonds, no matter howe expensive they get, and bonds can't turn into stocks, no matter how cheap they get. There's no migration between the two, so the spread has to shrink for the equity risk premium to shrink. But there's migration between value stocks and growth stocks, so things are not nearly that simple.

Does that help?
Last edited by backpacker on Tue Aug 23, 2016 3:43 pm, edited 1 time in total.

User avatar
Topic Author
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: Better growth? Or on how the value premium ends

Post by backpacker » Tue Aug 23, 2016 3:10 pm

backpacker wrote:I would love to be shown wrong by an argument that displayed real understanding of the issues rather than a lame appeal to authority. Do you have one?
larryswedroe wrote:I'm done, nothing more I can say except your arguments are incorrect. I've tried my best to explain why. so will not be responding any longer. But it's interesting to think that you understand the math and the financial issues better than the leading financial economists...
I guess the answer is no. :D

There are lots of interesting questions here Larry and I'm sorry you're not intellectually curious enough to see them. You're missing out!

Suppose you're right and that spreads are a good way to measure the value premium. Well why is that? That spreads shrink as the value premium shrinks is not a mathematical fact. That's what my examples show. If they do shrink as the premium shrinks, that follows from interesting empirical facts about how stocks are priced. Well, which ones? And what reason do we have to believe them? Maybe the smart financial types you're fond of know what those facts are and, if so, good for them. But I want to understand what those facts are and not just be told that someone else does.
Last edited by backpacker on Tue Aug 23, 2016 3:35 pm, edited 1 time in total.

User avatar
nedsaid
Posts: 12668
Joined: Fri Nov 23, 2012 12:33 pm

Re: Better growth? Or on how the value premium ends

Post by nedsaid » Tue Aug 23, 2016 3:32 pm

backpacker wrote:
nedsaid wrote: Looks like Larry is right here. These are all funds that I own except for Vanguard Value Index and DFA. All except for my Plain Old Boring Large Value fund had a higher volatility measure than the S&P 500.
An interesting observation here nedsaid. One question that I don't know the answer to is how to distinguish different kinds of risk. Whenever you carve off a chunk of the market (like value), you should expect that chunk to have more volatility than the market as a whole. The volatility of an energy sector fund will be higher than the volatility of the S&P for instance, but that doesn't show that energy companies have more risk than average. The energy fund has more volatility only because it is less diversified, not because energy companies are inherently riskier. The question is how much value funds are like that. Are they riskier just because they are less diversified than the total market?
That is a good question, Backpacker. Not sure that I have an answer any better than yours.

I have always thought of Value as being LESS risky than the market as a whole because they have less expectation built into their price than the broad market and even less than for Growth stocks. A popular growth stock that has a big earnings miss can just crater. If a Value stock has an earnings miss, few really care. But when I looked at the volatility stats, Larry was right, Value was more volatile than the market over 15 years. What I should have done was a comparison of the volatility of Value vs. Growth, but pretty much I assume that the S&P 500 is just loaded with all the large and popular growth funds.

My guess is that the fundamental risk of value stocks outweighs what I perceive to be their lower expectations risk. That would explain why Value stocks are more volatile. So you might say that Value stocks are crummy companies but not as crummy as the market thinks.

But you raise a point, that perhaps any subset of the market is riskier than the market as a whole. If you randomly sampled stocks from the broad market, probably 200-300 would be a large enough sample to mirror market returns. The question is, would volatility of the sample also mirror the market? Does adding more stocks to a sample always reduce volatility? My off the top scientific wild guess is that after maybe 400 you would have very little gain in return or reduced volatility after that. Dr. Bernstein seems to think the more, the better.

So lets compare the S&P 500 Index to the Total Stock Market. LOL!!! LOL!!! LOL!!! I am checking the numbers and my thesis is blown right out of the water. Total Market had a 6.63% annual return and 15.07 standard deviation. The S&P 500 had a 6.08% return but a 14.71 standard deviation. The Total Market has about 3,300 stocks and the S&P 500 has, well, 500. Adding more stocks did not reduce volatility, so there is something else at work here. Indeed, the Total Market outperformed the S&P 500 by 55 basis points a year. So you see the small-cap effect here.

Larry would tell you that the extra return and the extra volatility of the Total Market over the S&P 500 was due to factors and not the size of the sample. Just eyeballing here would tell me that Larry is correct. (Darn it!) So it is the characteristics of the stocks that affect their volatility and not the size of the sample.
A fool and his money are good for business.

bjr89
Posts: 149
Joined: Sat Apr 23, 2016 8:15 am

Re: Better growth? Or on how the value premium ends

Post by bjr89 » Tue Aug 23, 2016 3:54 pm

-----
Last edited by bjr89 on Fri Nov 16, 2018 7:36 pm, edited 3 times in total.

User avatar
JoMoney
Posts: 7672
Joined: Tue Jul 23, 2013 5:31 am

Re: Better growth? Or on how the value premium ends

Post by JoMoney » Tue Aug 23, 2016 3:54 pm

backpacker wrote:
nedsaid wrote: Looks like Larry is right here. These are all funds that I own except for Vanguard Value Index and DFA. All except for my Plain Old Boring Large Value fund had a higher volatility measure than the S&P 500.
An interesting observation here nedsaid. One question that I don't know the answer to is how to distinguish different kinds of risk. Whenever you carve off a chunk of the market (like value), you should expect that chunk to have more volatility than the market as a whole. The volatility of an energy sector fund will be higher than the volatility of the S&P for instance, but that doesn't show that energy companies have more risk than average. The energy fund has more volatility only because it is less diversified, not because energy companies are inherently riskier. The question is how much value funds are like that. Are they riskier just because they are less diversified than the total market?
Having larger stocks seems to reduce the standard deviation.
VFINX (Vanguard S&P 500) has a lower standard deviation than VTSMX (Total Stock Market)
OEF (iShares S&P 100) has a lower standard deviation than the S&P 500

VIVAX (Vanguard's Value Index) currently has a median market cap of $85.1 Billion compared to VIGRX (Growth Index fund) $71.5 Billion, and VFINX (Vanguard 500) $85.0 Billion.

Based on relative size, and the tendency for larger-caps to have lower standard deviation, I would expect the Value fund might continue to have slightly lower SD. Vanguards High-Dividend index VYM has had an even larger median market cap, and lower standard deviation over it's existence...
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

User avatar
nedsaid
Posts: 12668
Joined: Fri Nov 23, 2012 12:33 pm

Re: Better growth? Or on how the value premium ends

Post by nedsaid » Tue Aug 23, 2016 4:11 pm

JoMoney wrote:
backpacker wrote:
nedsaid wrote: Looks like Larry is right here. These are all funds that I own except for Vanguard Value Index and DFA. All except for my Plain Old Boring Large Value fund had a higher volatility measure than the S&P 500.
An interesting observation here nedsaid. One question that I don't know the answer to is how to distinguish different kinds of risk. Whenever you carve off a chunk of the market (like value), you should expect that chunk to have more volatility than the market as a whole. The volatility of an energy sector fund will be higher than the volatility of the S&P for instance, but that doesn't show that energy companies have more risk than average. The energy fund has more volatility only because it is less diversified, not because energy companies are inherently riskier. The question is how much value funds are like that. Are they riskier just because they are less diversified than the total market?
Having larger stocks seems to reduce the standard deviation.
VFINX (Vanguard S&P 500) has a lower standard deviation than VTSMX (Total Stock Market)
OEF (iShares S&P 100) has a lower standard deviation than the S&P 500

VIVAX (Vanguard's Value Index) currently has a median market cap of $85.1 Billion compared to VIGRX (Growth Index fund) $71.5 Billion, and VFINX (Vanguard 500) $85.0 Billion.

Based on relative size, and the tendency for larger-caps to have lower standard deviation, I would expect the Value fund might continue to have slightly lower SD. Vanguards High-Dividend index VYM has had an even larger median market cap, and lower standard deviation over it's existence...
JoMoney, you made Larry's factor argument. Yes, Large Stocks are less volatile than Small Stocks. But oddly enough, Large Growth has a smaller median market cap than Large Value but Large Growth is actually less volatile than Large Value. Large Value beats Large Growth in performance but not by much. You can see that there is not only a size factor but also a value factor as well.

So we are seeing two sets of characteristics at work here, the size of the stock and its value characteristics. In the case of High Dividend, the lower volatility would be due to not only to larger market cap but also steadier earnings. You can fake a dividend for a short time but ultimately you need the earnings to support them. Wall Street loves consistency!

But of course, the numbers I am using are over the last 15 years. Pick different time periods and the results might be different. Also what I am doing is eyeballing and mentally extrapolating but my process is not scientific. I will leave it to the quant and stats guys to put meat on the bones. I can barely post a chart! From what I can see, it appears Larry is right.
A fool and his money are good for business.

User avatar
JoMoney
Posts: 7672
Joined: Tue Jul 23, 2013 5:31 am

Re: Better growth? Or on how the value premium ends

Post by JoMoney » Tue Aug 23, 2016 4:16 pm

Oddly, despite Vanguard CRSP value/growth indexes, the Russell 1000 Growth vs Value has a larger market cap on the growth side.
https://personal.vanguard.com/us/funds/ ... =INT#tab=2

https://personal.vanguard.com/us/funds/ ... =INT#tab=2
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham

larryswedroe
Posts: 15989
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Better growth? Or on how the value premium ends

Post by larryswedroe » Tue Aug 23, 2016 4:16 pm

Few things
First, as I stated as Jim Davis of DFA showed in a paper the value premium spread predicts the value premium, larger spreads predict higher returns
Second, not hanging around because I've explained why backpacker is incorrect and cannot explain it better. If you believe the value story is risk then there must continue to be a value spread, period. End of discussion. Third, if you think it's behavioral then the spread narrows as people arb away the mispricing, period, end of discussion.
So whatever you believe the value premium is predicted by the size of the spread.
Larry

bjr89
Posts: 149
Joined: Sat Apr 23, 2016 8:15 am

Re: Better growth? Or on how the value premium ends

Post by bjr89 » Tue Aug 23, 2016 4:18 pm

-----
Last edited by bjr89 on Fri Nov 16, 2018 9:00 pm, edited 1 time in total.

bjr89
Posts: 149
Joined: Sat Apr 23, 2016 8:15 am

Re: Better growth? Or on how the value premium ends

Post by bjr89 » Tue Aug 23, 2016 4:20 pm

-----
Last edited by bjr89 on Fri Nov 16, 2018 9:00 pm, edited 1 time in total.

larryswedroe
Posts: 15989
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Better growth? Or on how the value premium ends

Post by larryswedroe » Tue Aug 23, 2016 5:18 pm

bjr
Ask why the companies are selling at 20 and 10 pes, especially if they earn the same. Could it be one is riskier than the other for whatever the reasons.
Now as I said one can concoct artificial single company examples perhaps, but not for markets.
Remember value stocks are not determined by say P/E which is just one metric. So you can have a higher P/E simply due to higher growth in earnings. But then you would have say differences in p/b ratios and other metrics. Remember typical value companies have simple risk characteristics of more financial and operating leverage and more volatile earnings, etc. They are then clearly riskier and thus require risk premiums and thus have higher costs of capital and higher expected returns.
As I said, the there is NO WAY you can have either a risk story or behavioral story and not have premium shrink without spread shrinking.Larry

User avatar
nedsaid
Posts: 12668
Joined: Fri Nov 23, 2012 12:33 pm

Re: Better growth? Or on how the value premium ends

Post by nedsaid » Tue Aug 23, 2016 5:29 pm

JoMoney wrote:Oddly, despite Vanguard CRSP value/growth indexes, the Russell 1000 Growth vs Value has a larger market cap on the growth side.
https://personal.vanguard.com/us/funds/ ... =INT#tab=2

https://personal.vanguard.com/us/funds/ ... =INT#tab=2
If my theory doesn't work, just blame the indexes! Thanks.
A fool and his money are good for business.

Post Reply