Author Topic: Efficient market hypothesis, index investing and value investing  (Read 14778 times)

k9

  • Stubble
  • **
  • Posts: 241
  • Age: 39
Efficient market hypothesis, index investing and value investing
« on: December 07, 2012, 07:43:21 AM »
Many here believe in the efficient market hypothesis which states, among other things, that an individual investor cannot beat the market indices except by pure luck (and, on average, lucky winners will compensate unlucky losers).

In a way, I think that's true, stock picking is a lot of work for a very weak reward and actively managed funds actually don't beat, on average, their reference index, so you are paying fees for nothing. I've already told I'm a permanent portfolio addict, the PP is based on that theory, and I think the safest bet you can make is investing to reach the average, with index funds and an asset allocation that fits your needs.

However, I also think it's possible to beat the averages without just pure luck. I really think buying stocks on a bargain (wrt their intrinsic, fundamental value) leads on average to better results than just buying, well, all the stocks on the market. This has been proved through time by many studies, and many very successful investors are value investors (W. Buffet to name the most famous one). Don't talk about pure luck, a lot of students of Ben Graham have been very successful using the same underlying theories.

Finally, I think index investing has its drawbacks too. If you hold a S&P tracker, you are invested in Apple much more than any other stock precisely BECAUSE Apple is expensive. So you buy stock A instead of stock B just because A is more expensive than B : how the hell could it be the best investing strategy available on earth ?

Don't get me wrong, I really do think that index investing is the best choice for most of the money individual investors put on stocks, but I really wouldn't say that's because of hypothetic efficient markets (but because it's easier and a lot less stressful and it is a lot easier to keep disciplined).

As for me, I have currently about 15-20% of my invested money in individually chosen value stocks bought with a good margin of safety, all other money is in a PP.

k9

  • Stubble
  • **
  • Posts: 241
  • Age: 39
Re: Efficient market hypothesis, index investing and value investing
« Reply #1 on: December 07, 2012, 07:51:36 AM »
Another problem with index investing : imagine stock #501 skyrocketed recently, so it takes the place of stock #500 which is now excluded from SP500. So lots of institutional investors (mainly tracker vendors such as Vanguard) are FORCED to sell it massively as it does not fit in the tracker anymore. Many sellers, so the price gets down.

But the fundamental business of former #500 did not change at all : there was no bad news associated with it. If the index had been called SP600, #500's price wouldn't have changed that much. How could that even be compatible with EMH ? Stock #500 sounds like an oversold stock, how couldn't it be a bargain now ?

Karl

  • 5 O'Clock Shadow
  • *
  • Posts: 64
Re: Efficient market hypothesis, index investing and value investing
« Reply #2 on: December 07, 2012, 08:54:33 AM »
OK, value stocks have out-performed growth stocks over prolonged periods of time (point accepted).  However, I do not know of any evidence that most investors (including you and me) can do better than to purchase a value index.  I admit that some people have done well as individual investors in value stocks.  However, I also admit that some people can play sports well enough to make a living at it.  I don't know of any reason to assume that I can perform professionally in either area (or well enough to simply justify the investment of my time).  I know I can improve my own financial position by working to reduce expenses, become more expert at my current job to try to boost my income, and I can spend my time to become a better father/spouse.  I hope you come out ahead on your personal investments, but you might also consider if you simply do it because you find it enjoyable. 

k9

  • Stubble
  • **
  • Posts: 241
  • Age: 39
Re: Efficient market hypothesis, index investing and value investing
« Reply #3 on: December 07, 2012, 09:12:29 AM »
The problem with value indices is that, once included in an index, a stock is often a lot less of a bargain than before :( because so many tracker vendors are going to buy it. And becoming an individual superstar value investor is, indeed, impossible for aalmost everybody.

However I do think it's possible to make a virtual value index buy buying small caps (those no fund will ever buy because it's not possible to invest the amounts of money they have to) that check a given value criterion, for example, the 20 very-small-caps that have the smallest P/E ratio in the US. That's an index too, after all, but instead of buying the biggest companies, you buy the ones whose earnings are the cheapest. And it has been proved, on the long term, to give a better return (with less risk), provided you stick with it (and that's the hard part). The "dogs of the dow" is a value investing strategy that does not require the investor to do any work at all (beside sorting the dow jones companies according to their dividend yield and buying the 10 first). It is not as "value" as other strategies, and it won't provide amazingly good returns, but that's the kind of thing anybody can implement.

COguy

  • 5 O'Clock Shadow
  • *
  • Posts: 99
  • Location: Longmont, Colorado
Re: Efficient market hypothesis, index investing and value investing
« Reply #4 on: December 07, 2012, 09:55:25 AM »
Quote
Another problem with index investing : imagine stock #501 skyrocketed recently, so it takes the place of stock #500 which is now excluded from SP500. So lots of institutional investors (mainly tracker vendors such as Vanguard) are FORCED to sell it massively as it does not fit in the tracker anymore. Many sellers, so the price gets down.

This is one of the reasons why holding Total Market Funds can be preferable to S and P 500 trackers and the like. 


chucklesmcgee

  • Pencil Stache
  • ****
  • Posts: 613
Re: Efficient market hypothesis, index investing and value investing
« Reply #5 on: December 07, 2012, 10:33:25 AM »
However I do think it's possible to make a virtual value index buy buying small caps (those no fund will ever buy because it's not possible to invest the amounts of money they have to) that check a given value criterion, for example, the 20 very-small-caps that have the smallest P/E ratio in the US. That's an index too, after all, but instead of buying the biggest companies, you buy the ones whose earnings are the cheapest. And it has been proved, on the long term, to give a better return (with less risk), provided you stick with it (and that's the hard part). The "dogs of the dow" is a value investing strategy that does not require the investor to do any work at all (beside sorting the dow jones companies according to their dividend yield and buying the 10 first). It is not as "value" as other strategies, and it won't provide amazingly good returns, but that's the kind of thing anybody can implement.


So there is at least one dogs of the dow ETF:
http://finance.yahoo.com/q/pr?s=SDOG


"for the 20 years from 1992 to 2011... the Dogs of the Dow matched the average annual total return of the Dow (10.8%) and outperformed the S&P 500 (9.6%) as reported by the Dogs of the Dow website located at dogsofthedow.com. The Small Dogs of the Dow, which are the five lowest priced Dogs of the Dow, outperformed both the Dow and S&P 500 with an average annual total return of 12.6%.[1] When each individual year is reviewed it is clear that both the Dogs of the Dow and Small Dogs of the Dow did not outperform each and every year. In fact, the Dogs of the Dow and Small Dogs of the Dow struggled to keep up with the Dow during latter stages of the dot-com boom (1998 and 1999) as well as during the financial crisis (2007-2009).[2] This suggests that an investor would be best served by viewing this as a longer-term strategy by giving this portfolio of stocks time to recover in case of a rare but extreme economic event (e.g. dot-com boom, financial crisis)"

So I don't really see this as an especially spectacular strategy. Had I put my money into a Vanguard total stock index (Dow), dogs of the dow would have matched my returns I'd have come out ahead because of the lower fees. And since the Dogs is necessarily a smaller set of holdings, they're more volatile. I'd need to see a statistical analysis, but I really wonder if this set of 20, 40 or however many stocks has really significantly deviated from the returns you'd expect from 20-40 randomly selected stocks. In the end it doesn't support the idea that a typical investor would expect to outperform the market better than chance.

Quote
I really think buying stocks on a bargain (wrt their intrinsic, fundamental value) leads on average to better results than just buying, well, all the stocks on the market.

"Really thinking" something doesn't make it true! Can you define intrinsic, fundamental value of a stock for us in a quantitative way we could actually test it?  And if it's just a super-magic-really-good "investor gut feel" to buy a certain stock that can't be quantified, surely some set of people have this talent and knack, right? A lot of those people are probably going to become fund managers and stock gurus. Hmm, but actively managed funds can't beat the index. And stock gurus, when they're actually making enough explicit stock picks, can be shown to do better than chance. So how can both be true?

Quote
This has been proved through time by many studies, and many very successful investors are value investors (W. Buffet to name the most famous one).

Citation showing that value investments consistently outperform the markets needed. That some successful investors are value investors is not good evidence that individuals can outperform the markets by chance. Buffet has enough capital to actually influence companies, decisions and management- if his team's better at managing a company than the current staff, obviously he can get better returns, especially considering that he'll have a good deal of inside information not available to the average investor. At this point too, his investments are a self-fulfilling prophecy. When news broke that he had invested $5 billion in Bank of America, the stock soared 25% in a single day, netting him an easy billion. That unfortunately means that even if you had Buffet's strategy and knack in your brain, you couldn't match Buffet's returns without his information, fame and pile of cash.

Quote
Don't talk about pure luck, a lot of students of Ben Graham have been very successful using the same underlying theories

Mhm. That's exactly what pure luck would suggest. Have all investors applying Graham's theories consistently outperformed the markets? Has the average Graham investor outperformed the market over a decent timespan?  That you can point to a dozen or so individuals who have been successful and claimed to have used Graham's theories isn't good evidence they work any better than an index fund. I could get 1000 monkeys to pick a few dozen stocks and half would outperform the market and just a fraction would have negative returns

Where's the evidence that your value-investors can beat my monkeys?

KingCoin

  • Pencil Stache
  • ****
  • Posts: 783
  • Location: Manhattan
  • Achieved FI @ 30
Re: Efficient market hypothesis, index investing and value investing
« Reply #6 on: December 07, 2012, 11:23:44 AM »
Everything chucklesmcgee said +1

Finally, I think index investing has its drawbacks too. If you hold a S&P tracker, you are invested in Apple much more than any other stock precisely BECAUSE Apple is expensive. So you buy stock A instead of stock B just because A is more expensive than B : how the hell could it be the best investing strategy available on earth ?

A market cap weighted index makes much more sense than an equal weight index as far as I'm concerned.  Just because Apple's market cap is high, doesn't mean it's "expensive". An equal weight index invests a disproportionate amount in smaller cap companies. These tend to outperform larger cap companies over the long term, but with higher average volatility (which is consistent with the efficient investment frontier)*. Imagine a simplified 2 company economy. One company is a 2 trillion behemoth and the other is a 20 million microcap. Would it make sense to invest an equal amount of your money in each? Of course not.

You could make a reasonable argument that the index should be weighted by earnings rather than market cap. There are ETF and mutual funds that take this approach if you're so inclined. These will naturally invest more heavily in lower P/E companies; potentially interesting if you do some research and decide that's a strategy you want to pursue.

* http://seekingalpha.com/article/266279-have-small-caps-peaked
« Last Edit: December 07, 2012, 11:31:56 AM by KingCoin »

smedleyb

  • Bristles
  • ***
  • Posts: 434
Re: Efficient market hypothesis, index investing and value investing
« Reply #7 on: December 07, 2012, 12:35:31 PM »
As for me, I have currently about 15-20% of my invested money in individually chosen value stocks bought with a good margin of safety, all other money is in a PP.

Which stocks do you own?

MooreBonds

  • Stubble
  • **
  • Posts: 193
  • Age: 42
  • Location: St. Louis, MO
Re: Efficient market hypothesis, index investing and value investing
« Reply #8 on: December 07, 2012, 06:16:48 PM »
I really think buying stocks on a bargain (wrt their intrinsic, fundamental value) leads on average to better results than just buying, well, all the stocks on the market.

I agree with you, and also try to be more of a value investor (on average)....however, even value investing is subject to your own individual assumptions -just like growth companies - because there's no one single "intrinsic worth" value of a company or entry on a balance sheet. Your discount rate, projections of future earnings, value of assets, etc. will be different than mine. As such, everyone has different intrinsic worth values of each stock. So just as everyone has different estimates of growth stocks' future earnings, the same applies to what people value for future earnings and assets of the intrinsic worth of a company.


Of course, if you are dealing with a situation of a company's reasonable liquidation value or net cash being greater than its market value, that's the extreme case. But for most others, it's all a question of what each investor's assumption is on value, and what the next/alternative investment is offering for potential value.

k9

  • Stubble
  • **
  • Posts: 241
  • Age: 39
Re: Efficient market hypothesis, index investing and value investing
« Reply #9 on: December 08, 2012, 05:22:37 AM »
So I don't really see this as an especially spectacular strategy. Had I put my money into a Vanguard total stock index (Dow), dogs of the dow would have matched my returns I'd have come out ahead because of the lower fees. And since the Dogs is necessarily a smaller set of holdings, they're more volatile. I'd need to see a statistical analysis, but I really wonder if this set of 20, 40 or however many stocks has really significantly deviated from the returns you'd expect from 20-40 randomly selected stocks. In the end it doesn't support the idea that a typical investor would expect to outperform the market better than chance.
O'Shaugnessy in "what works on wall street" computed the result of "dogs of the dow" when compared to SP500 for a period between 1929 to 2004.

SP500 : 9.7%/year, sharpe ratio 0.38, min return -43.34%, max return 53.99%
dogs : 12.24%/year, sharpe ratio 0.49, min return -48.88%, max return 66.73%

That's a really big difference for such a long period. Fees must be included and will lower the return, but selling&buying between 0 and 10 securities per year isn't that expensive. Anyway, that strategy is far from being my favorite value strategy, but its the most famous one.

Quote
Can you define intrinsic, fundamental value of a stock for us in a quantitative way we could actually test it?
You could test something as simple as price-to-earnings ratio < 10. Or current price < 67% * tangible book value. These simple metrics have been advised by Ben Graham at least as early as the 1970s (maybe even earlier, I can't be sure) and have been checked in the following decades (so that's not just a happy result of datamining past returns). You won't become Warren Buffet just that way, but in the past (and I can't see why it won't happen again) you would have beaten the average by a few points each year.

Quote
And if it's just a super-magic-really-good "investor gut feel" to buy a certain stock that can't be quantified, surely some set of people have this talent and knack, right? A lot of those people are probably going to become fund managers and stock gurus. Hmm, but actively managed funds can't beat the index. And stock gurus, when they're actually making enough explicit stock picks, can be shown to do better than chance. So how can both be true?
Well, there's Berkshire Hattaway at least.

There are a few problems with value fund managers, actually. First, they can't buy really small caps, because the amounts of money they deal with don't let them buy enough stocks at the current price. Second, investors don't like their money to underperform the market two years or more in a row. But value investing is not a steady road ; it's a very long term, very volatile trip. You can have very bad performance for a few years but skyrocket the year after. Well, the problem is customers of your fund will hate you on bad years, and most of them will leave you then. Everybody said Buffet was "dead" in 1998-1999 because he was underperforming the dot.com investors.

Quote
Citation showing that value investments consistently outperform the markets needed.
Have a look at the tweedy, browne study (http://www.tweedy.com/resources/library_docs/papers/WhatHasWorkedFundVersionWeb.pdf) that has tested various "value" metrics trough time. Joseph Piotroski's work are interesting too and point to many academic papers on that subject (http://www.chicagobooth.edu/faculty/selectedpapers/sp84.pdf)

A market cap weighted index makes much more sense than an equal weight index as far as I'm concerned.  Just because Apple's market cap is high, doesn't mean it's "expensive".
The problem is, if it gets significantly higher today because of a good news, its weight in the index will grow and tracker vendors will have to buy some more, thus driving it even a little higher. That doesn't sound efficient to me.

Index investing is shown as an obvious, no-brainer solution to invest in the markets. Well, it's not that obvious, it's a strategy per se : a momentum strategy. The more the price of a security grows, the more you buy it. That's not "natural", that's a strategy in itself. But as you point, equal weighting is not "natural" either, because small companies that don't have effects on the overall economy will impact your portfolio. There is no perfect, neutral, natural way to invest in the markets. All imply you follow a strategy. Index investing is one of them, no more, no less. It is in now way inherently better. It's just easier.

Smedleyb, I'm invested in French small-caps & micro-caps.

MooreBonds, you're right, that's why I prefer sticking to strategies that have been documented and checked in the past.

rjack

  • 5 O'Clock Shadow
  • *
  • Posts: 90
  • Age: 59
  • Location: Philadelphia PA
  • I'm retired!
Re: Efficient market hypothesis, index investing and value investing
« Reply #10 on: December 08, 2012, 06:06:02 AM »
Where's the evidence that your value-investors can beat my monkeys?

You might want to look here:

http://www.bengrahaminvesting.ca/Research/Academic_Research/published_papers.htm


KingCoin

  • Pencil Stache
  • ****
  • Posts: 783
  • Location: Manhattan
  • Achieved FI @ 30
Re: Efficient market hypothesis, index investing and value investing
« Reply #11 on: December 08, 2012, 11:27:24 AM »
The problem is, if it gets significantly higher today because of a good news, its weight in the index will grow and tracker vendors will have to buy some more, thus driving it even a little higher. That doesn't sound efficient to me.

As the price rises, it will naturally become a larger part of the portfolio. Why would the manager have to buy more shares?

On the point of stocks getting a boost or selling off because of a new index inclusion or exclusion, there are countless hedge funds who attempt to arb this effect. Unfortunately, there's not much money to be made there.

sheepstache

  • Handlebar Stache
  • *****
  • Posts: 2424
Re: Efficient market hypothesis, index investing and value investing
« Reply #12 on: December 08, 2012, 02:14:44 PM »
Where's the evidence that your value-investors can beat my monkeys?

You might want to look here:

http://www.bengrahaminvesting.ca/Research/Academic_Research/published_papers.htm

Dude that's like 60 papers.  Did you have one in mind?

grantmeaname

  • Magnum Stache
  • ******
  • Posts: 4811
  • Age: 27
  • Location: NYC
  • Cast me away from yesterday's things
Re: Efficient market hypothesis, index investing and value investing
« Reply #13 on: December 08, 2012, 10:59:03 PM »
Many here believe in the efficient market hypothesis which states, among other things, that an individual investor cannot beat the market indices except by pure luck (and, on average, lucky winners will compensate unlucky losers).
Yay, another nerdy investment post!

Quote
This has been proved through time by many studies, and many very successful investors are value investors (W. Buffet to name the most famous one). Don't talk about pure luck, a lot of students of Ben Graham have been very successful using the same underlying theories.
Value investing is one of the biggest techniques active managers use. Doesn't this contradict what you're saying above? Also, what studies? I've never seen something saying that active managers using technical analysis can't beat the indices but that active managers using fundamental analysis can.

Quote
Finally, I think index investing has its drawbacks too. If you hold a S&P tracker, you are invested in Apple much more than any other stock precisely BECAUSE Apple is expensive. So you buy stock A instead of stock B just because A is more expensive than B : how the hell could it be the best investing strategy available on earth ?
Apple is a much bigger company than Dollar General or Monster Beverage or any of the other little companies in the S&P 500. If you want your results to mimic the performance of the economy as a whole, it's important that you hold companies in proportion to their size. If you owned 1% of Apple and an equal amount of Monster Beverage, you'd be a majority shareholder all on your own.

Another problem with index investing : imagine stock #501 skyrocketed recently, so it takes the place of stock #500 which is now excluded from SP500. So lots of institutional investors (mainly tracker vendors such as Vanguard) are FORCED to sell it massively as it does not fit in the tracker anymore. Many sellers, so the price gets down.

But the fundamental business of former #500 did not change at all : there was no bad news associated with it. If the index had been called SP600, #500's price wouldn't have changed that much. How could that even be compatible with EMH ? Stock #500 sounds like an oversold stock, how couldn't it be a bargain now?
That is a problem with index funds tracking some indices. However, there are a couple caveats that help minimize it. First, the S&P 500 is not the Fortune 500, and not just the biggest 500 companies in the US. They're hand selected by a totally arbitrary system. Joe the Contrarian Mutual Fund Speculator can't watch stock 501's market cap and know when it passes stock 500's market cap that the two will be exchanged. Then, when something is added or removed from the benchmark, funds don't announce when they're selling off and buying new, if they do so at all. That's the last, major point: Imagine you were a company managing index funds based on the Russell 1000 (1000 biggest companies) and Russell 2000 (stocks 1001-3000). When stock 1001 got promoted to 1000, you could just trade the shares between your two funds, especially if they had similar amounts invested in them (With the popularity of S&P 500 funds like SPY and VFINX in particular, that may not be entirely the case, because they may have more assets invested in them than the companies' other funds, but it still does serve to minimize the effect.)

grantmeaname

  • Magnum Stache
  • ******
  • Posts: 4811
  • Age: 27
  • Location: NYC
  • Cast me away from yesterday's things
Re: Efficient market hypothesis, index investing and value investing
« Reply #14 on: December 08, 2012, 11:26:56 PM »
Well, there's Berkshire Hattaway at least.
Warren Buffet is not just an investor. Compared to you, I, insurance companies, pension funds, and most everyone else just investing, Buffett has the ability to divert profits from businesses to other investments to provide "free" financing to new acquistions. He avoids the taxes that we pay on receiving dividends by not paying dividends. He controls the management of the companies he buys by buying a significant stake in them. He's had close friendships with the managers and chief owners of acquired companies, leading to SEC probes about inside information and collusion. He's got the sheer scale of capital needed to play and do well in currency futures. Point being, Buffett didn't just make his millions by studious application of the Graham rules.

Quote
There are a few problems with value fund managers, actually. First, they can't buy really small caps, because the amounts of money they deal with don't let them buy enough stocks at the current price.
You keep saying that, but I'm seeing micro-cap ETFs like WMCR holding thousands of pitifully tiny stocks. Additionally, even the Russell 3000 alone covers all but the bottom 2% of the market, and the Russell 1000 covers all but the bottom 10%. The big stocks are explaining 90% of the markets' total capitalization and 90% of the earnings and returns, so at most the small caps could be an amusing asterisk on the results, and that only if WMCR and its ilk didn't exist.

Quote
Second, investors don't like their money to underperform the market two years or more in a row. But value investing is not a steady road ; it's a very long term, very volatile trip. You can have very bad performance for a few years but skyrocket the year after. Well, the problem is customers of your fund will hate you on bad years, and most of them will leave you then. Everybody said Buffet was "dead" in 1998-1999 because he was underperforming the dot.com investors.
That doesn't change the performance of the fund, that changes the total amount under management. It's trivial enough to calculate the hypothetical growth of a single investment, and when the funds are listing their performance results or you're reading a page about them on Morningstar or a similar service, you're not reading the changes in total funds under management at the firm, but the effects on a constant amount of capital from the fund's activities only.



Where's the evidence that your value-investors can beat my monkeys?
You might want to look here:
http://www.bengrahaminvesting.ca/Research/Academic_Research/published_papers.htm
Dude that's like 60 papers.  Did you have one in mind?
Rjack, that's just a library. It cites Eugene Fama, the die-hard ideologue of a finance professor who invented the efficient market hypothesis, for example. If you had a study in mind, it would be better to point to it directly.

rjack

  • 5 O'Clock Shadow
  • *
  • Posts: 90
  • Age: 59
  • Location: Philadelphia PA
  • I'm retired!
Re: Efficient market hypothesis, index investing and value investing
« Reply #15 on: December 10, 2012, 01:56:42 PM »
Where's the evidence that your value-investors can beat my monkeys?

You might want to look here:

http://www.bengrahaminvesting.ca/Research/Academic_Research/published_papers.htm

Dude that's like 60 papers.  Did you have one in mind?

Wow...there are lots of useful research papers since it is The Ben Graham Center for Value Investing. Anyway, try:

http://www.bengrahaminvesting.ca/Outreach/Symposium/Papers/The_Performance_Pervasiveness_and_Determinants_of_Value_Premium_in_Different_US_Exchanges.pdf

or

http://www.bengrahaminvesting.ca/Research/Papers/Athanassakos/Do_Value_Investors_Add_Value(short).pdf

Also, if you are not interested or too lazy to dig around research like this, then stock picking is not for you and you should stick with index investing. Personally, I mostly use index investing.

grantmeaname

  • Magnum Stache
  • ******
  • Posts: 4811
  • Age: 27
  • Location: NYC
  • Cast me away from yesterday's things
Re: Efficient market hypothesis, index investing and value investing
« Reply #16 on: December 10, 2012, 03:04:07 PM »
Good! Ad hominem instead of ideas, that's always great to see.

k9

  • Stubble
  • **
  • Posts: 241
  • Age: 39
Re: Efficient market hypothesis, index investing and value investing
« Reply #17 on: December 11, 2012, 02:03:15 AM »
Apple is a much bigger company than Dollar General or Monster Beverage or any of the other little companies in the S&P 500. If you want your results to mimic the performance of the economy as a whole, it's important that you hold companies in proportion to their size. If you owned 1% of Apple and an equal amount of Monster Beverage, you'd be a majority shareholder all on your own.

That's right, even split between companies is not necessarily better (it is probably worse, actually, as you point out) but I wanted to point out that index investing is not as neutral as one could think. It is a strategy per se, a momentum strategy to be precise. Which is ironic when you think that index investing is advised in all lazy portfolios (including the PP I like) and when the main strategy of lazy portfolios is "rebalance often, sell your winning assets to buy more of your losers, return to the mean, buy low sell high, doing the opposite is market timing, yada yada yada.

Index investing is not neutral and not better. It is just easier to implement, less stressful and often less expensive than other strategies.

COguy

  • 5 O'Clock Shadow
  • *
  • Posts: 99
  • Location: Longmont, Colorado
Re: Efficient market hypothesis, index investing and value investing
« Reply #18 on: December 11, 2012, 05:03:47 PM »
I wouldn't say "rebalance often" is advised by the lazy portfolios like the PP and such.  Generally, it involves re-balancing bands.  Or, if you follow a Boglehead approach it would be once a year.  I think I read up once that the effects of rebalancing more than once a year are pretty small.  Plus, it just adds more work in which case one may be able to focus their efforts better elsewhere

After I typed this I wonder why I got involved in this conversation.  Personally, I use a very Jim Collins like approach in my investing.  I figure unless I have a 7 figure portfolio, I am going to have a bitch of a time generating a decent hourly wage through investing.  I would rather do other things with my life. 

Some people love analyzing companies and to them I hope them nothing but success.  Especially the mustachian ones.

Crash87

  • Stubble
  • **
  • Posts: 165
  • Age: 32
  • Location: St. Louis
Re: Efficient market hypothesis, index investing and value investing
« Reply #19 on: December 13, 2012, 05:55:41 PM »
I highly suggest giving "The Intelligent Asset Allocator" on MMM's suggested reading list a read. The author is pro-value indexes and has what I consider a convincing arguement.

But I'm generally new to investing so my opinion is of little value.

k9

  • Stubble
  • **
  • Posts: 241
  • Age: 39
Re: Efficient market hypothesis, index investing and value investing
« Reply #20 on: December 14, 2012, 02:59:25 AM »
After I typed this I wonder why I got involved in this conversation.  Personally, I use a very Jim Collins like approach in my investing.  I figure unless I have a 7 figure portfolio, I am going to have a bitch of a time generating a decent hourly wage through investing.  I would rather do other things with my life. 

Some people love analyzing companies and to them I hope them nothing but success.  Especially the mustachian ones.
Yes, that's right, unless you're very rich or you think it's fun, stock picking is not really worth the time spent. I'm in the second category, though. But I wouldn't advise people who don't like it and don't invest millions to follow anything other than index investing with a sound asset allocation.

grantmeaname

  • Magnum Stache
  • ******
  • Posts: 4811
  • Age: 27
  • Location: NYC
  • Cast me away from yesterday's things
Re: Efficient market hypothesis, index investing and value investing
« Reply #21 on: December 14, 2012, 03:12:50 PM »
Why does being rich make it worth your time?

JohnGalt

  • Bristles
  • ***
  • Posts: 481
  • Age: 34
  • Location: TX
Re: Efficient market hypothesis, index investing and value investing
« Reply #22 on: December 14, 2012, 03:53:13 PM »
Why does being rich make it worth your time?

Earning an extra 1%/year over index investing is more valuable the more you have to invest.  If you spend an extra 400 hours / year to get that extra 1%, you make $2.5/hour doing on $100,000 but $25/hour on $1,000,000.

grantmeaname

  • Magnum Stache
  • ******
  • Posts: 4811
  • Age: 27
  • Location: NYC
  • Cast me away from yesterday's things
Re: Efficient market hypothesis, index investing and value investing
« Reply #23 on: December 14, 2012, 04:00:50 PM »
Yes, but to earn 1% a year over index investing you have to be able to earn 1% a year over index investing, something that active managers haven't ever shown they're able to do.

COguy

  • 5 O'Clock Shadow
  • *
  • Posts: 99
  • Location: Longmont, Colorado
Re: Efficient market hypothesis, index investing and value investing
« Reply #24 on: December 14, 2012, 04:59:03 PM »
Quote
Yes, but to earn 1% a year over index investing you have to be able to earn 1% a year over index investing, something that active managers haven't ever shown they're able to do.

Exactly,  to me I think the best course is to leave that 1 million dollars in some passively managed scheme and just go work 400 hours for 25$/hour and ensure that I actually get that extra $10000.

I don't know about the rest of you but, if I had a million dollars, I wouldn't want a job.  Especially one that has a less than certain payout.  To each their own I guess.

chucklesmcgee

  • Pencil Stache
  • ****
  • Posts: 613
Re: Efficient market hypothesis, index investing and value investing
« Reply #25 on: December 15, 2012, 04:26:32 PM »
Why does being rich make it worth your time?

Earning an extra 1%/year over index investing is more valuable the more you have to invest.  If you spend an extra 400 hours / year to get that extra 1%, you make $2.5/hour doing on $100,000 but $25/hour on $1,000,000.

Well then clearly active index fund managers in the world have an enormous incentive to outperform the markets, considering that their expense ratios can be greater than 1% and they're easily managing billions of dollars....oh but wait, they can't beat the markets above chance.  I guess they're just lazy and untalented like all the other really good stock pickers who can't be found and tested.

Richard3

  • Bristles
  • ***
  • Posts: 292
Re: Efficient market hypothesis, index investing and value investing
« Reply #26 on: December 15, 2012, 05:29:47 PM »
Isn't the reason most active funds don't beat the market because of their fees rather than because of bad stock picking?

k9

  • Stubble
  • **
  • Posts: 241
  • Age: 39
Re: Efficient market hypothesis, index investing and value investing
« Reply #27 on: December 17, 2012, 03:42:58 AM »
It's also because if they don't take risks, do like anyone else and don't beat the average, whatever that average is, they keep their job. If they take risks and win, well, they keep their jobs and win a few more customers. But if they take risks and fail, they're fired. What would you do in their situation ?

For time spent in active investing vs stock picking, I don't really agree either. It's not between 1 hr/day or no time at all. An approach like coffee can portfolio (chose a dozen of stocks wisely, i.e. sound businesses that are undervalued wrt, say, p/e value, buy them and keep them for 10 years) doesn't take a lot of time (say, one week every 10 years) and provides better results, on average, than index investing (as you don't have to pay any management fee at all beside broker fees). The dividend aristocrat approach (buy only blue chips that offer ever-growing dividends and hold them forever) works too, and has much less volatility. They don't fit well with regular rebalancing, though.

chucklesmcgee

  • Pencil Stache
  • ****
  • Posts: 613
Re: Efficient market hypothesis, index investing and value investing
« Reply #28 on: December 17, 2012, 07:06:14 PM »
An approach like coffee can portfolio (chose a dozen of stocks wisely, i.e. sound businesses that are undervalued wrt, say, p/e value, buy them and keep them for 10 years...provides better results, on average, than index investing (as you don't have to pay any management fee at all beside broker fees.

Citation needed. Easily testable, easily disprovable, back that claim up.

k9

  • Stubble
  • **
  • Posts: 241
  • Age: 39
Re: Efficient market hypothesis, index investing and value investing
« Reply #29 on: December 18, 2012, 11:48:55 AM »
The rationale for index investing is that you'll get average returns before fees, so we can take that as an hypothesis. That means any other 100% stock allocation will provide the same results as an index, on average and before broker fees. Thus a strategy that costs less than an index tracker buy&hold will win. With the coffee can, you buy&sell one stock per year, so depending on your broker&tracker fees, it's easy to see what is less expensive. For my broker & available trackers, the coffee can is clearly less expensive (0.3% for buying/selling, 0.2% for the tracker fees), but your results might differ.

Putting $10000 on a SP500 tracker for 40 years, selling it back after 40 years :
10 000 * 0.997 * 0.998^40 * 0.997 = 9175 $

Buying 10 randomly chosen stocks from the SP500 for 1000$ each, selling them after 10 years, buying 10 new, for 40 years (so, 4 buying/selling operations) :
10 000 * 0.997^(2*4) = 9762 $

I have not included the returns of the investments, but they are supposed to be the same because of the way indexing is supposed to work. The coffee can has clearly less fees. The real problem with stock picking is that people get in & out of the market way too often, thus paying a lot of fees to their broker and being out of the market on good days.

For actual results : you can read Kirby's paper about the coffee can portfolio, but that's just an example, not a formal proof. James Montier talks about it in the little book of behavioral investing, too.

COguy

  • 5 O'Clock Shadow
  • *
  • Posts: 99
  • Location: Longmont, Colorado
Re: Efficient market hypothesis, index investing and value investing
« Reply #30 on: December 18, 2012, 02:30:52 PM »
Lets pretend that you can actually pick 4 sets of 10 stocks that replicate some index.  What if instead we do this the Jim Collins way?  if we have $10,000 we can do 100% in VTSAX ER=0.06% and no transaction fees?

Putting $10000 on a VTSAX for 40 years, selling it back after 40 years :
10 000 * 0.9994^40 = $9762

Buying 10 randomly chosen stocks from the SP500 for 1000$ each, selling them after 10 years, buying 10 new, for 40 years (so, 4 buying/selling operations) :
10 000 * 0.997^(2*4) = $9762

Plus you have bid/ask spreads (close to negligible for large caps and technically exist in index funds as well)...and taxes if in a taxable account.

Then, what if some unknown event happens over a 10 year span and one of your stocks drops 100%?  I would say you will end up with substantial tracking error from an index.

I think you would be hard pressed to make an inexpensive method that is cheaper than all in VTSAX for 40 years.  It is even harder in a taxable account. 

Of course you could get better returns with your 10 stocks






KingCoin

  • Pencil Stache
  • ****
  • Posts: 783
  • Location: Manhattan
  • Achieved FI @ 30
Re: Efficient market hypothesis, index investing and value investing
« Reply #31 on: December 18, 2012, 04:44:23 PM »
Then, what if some unknown event happens over a 10 year span and one of your stocks drops 100%?  I would say you will end up with substantial tracking error from an index.

Right. Most authorities prescribe 20-30 stocks to be "diversified".

On the low end, that's $500 each. Most discount brokerages charge $4-20/per trade, so 1%+ is probably a more realistic trade cost assumption. That puts a Vanguard fund way ahead.

In any event, it's going to add up to minutia compared to the long term compounding effect of a ~7% return.

I read this summary of the coffee can portfolio:
http://lmcm.com/868718.pdf
It seemed a really long winded way of saying that active trading can be detrimental, but I didn't see any evidence that stashing 10 stocks is more interesting than buying the index for the long term. Furthermore, as you acquire more investable funds over the 10 year waiting period, how do you decide how to allocate those funds to the stocks in the coffee can?  Seems pointlessly complicated.

k9

  • Stubble
  • **
  • Posts: 241
  • Age: 39
Re: Efficient market hypothesis, index investing and value investing
« Reply #32 on: December 19, 2012, 06:08:08 AM »
Here in France you can't find trackers that are that cheap. Of course, 0.06%, that probably changes the overall result. Here, 0.35% is very cheap.

Yes I know a stock can make -100%, but it can also go +500% or more on the very long run (10 years). The goal is not to replicate the index but to have, on average, the same performance (meaning half investors will do better, half will do worse). You can go with 20 or 30 to get closer to it, though, but that's starting to get expensive.

As for adding funds vs not investing all at once, I can't see how hard it is. Every time you get new funds you buy a new stock or reinvest in one of yours. It's much easier than buying a house & finding someone to pay its rent, IMO.

KingCoin

  • Pencil Stache
  • ****
  • Posts: 783
  • Location: Manhattan
  • Achieved FI @ 30
Re: Efficient market hypothesis, index investing and value investing
« Reply #33 on: December 19, 2012, 10:22:56 AM »
As for adding funds vs not investing all at once, I can't see how hard it is. Every time you get new funds you buy a new stock or reinvest in one of yours.

The problem with this is that you'll then have to constantly pick new stocks, or make decisions about which of your coffee can stocks to invest more in. During this frequent stock selection process, you're naturally subject to various psychological forces and cognitive biases. This is anathema to the spirit of investing in both indices and the "coffee can portfolio". Without any obvious advantages (the management fees saved are small at best), it's not clear why you'd choose stock picking (even if it's ostensibly buy and hold) over indexing.

k9

  • Stubble
  • **
  • Posts: 241
  • Age: 39
Re: Efficient market hypothesis, index investing and value investing
« Reply #34 on: December 20, 2012, 03:17:15 AM »
In the end I think it's a matter of personal choice. I practice both (index investing & stock picking) and they really feel very different. The former feels like saving money into an account that yields a lot but is very volatile (easy to do, just give them the money and look it grow - or not) while stock picking feels like really buying portions of a business I have chosen and that seems both cheap and able to produce a lot of cashflow on the long run. It's quite the same difference as between investing in REITs vs buying a house. The latter involves much more work (and sometimes, risk) but some investors can't imagine investing in REITs, because of fees and because they want to make their own choices.

k9

  • Stubble
  • **
  • Posts: 241
  • Age: 39
Re: Efficient market hypothesis, index investing and value investing
« Reply #35 on: February 03, 2013, 10:55:42 AM »
Hello all, I'm back on this topic because I just read an (old) article by James Montier about index investing that I found interesting and very close to what was discussed in this thread.

http://behaviouralinvesting.blogspot.fr/2007/09/something-boglehead-wouldnt-want-you-to.html

Quote
[...]on one of the misnomers of investing, that index investing is passive. This simply isn't true. Many indices are relatively actively managed. In fact most indices are really momentum players effectively adding stocks that have done well and deleting stocks that have done badly. This raises the question as to whether this 'active' element adds or destroys value.

[...]an investor would be 2.2% better off in year one if they ignored the index changes, this rises to 17% in year five! So a buy and hold strategy seems to generate substantially higher returns for investors

[...]The bottom line appears to be that index investing is often very far from passive. The rules of index construction appear to destroy value.

Basically it says that index investing is not really passive. Indexes are actually actively managed, with stocks going in & out of it. Index investing is a momentum strategy (that's very close to what I said earlier in this post, investors think they are market-neutral when buying and index, but they're not) where you buy winners and sell losers, and it has yet to be proven that momentum investing is better than real passive, buy&hold investing.

Well, he actually points to an article showing that buying&holding the securities from the S&P or Russel 2000 on a given year and never selling them provided better results than following the index's rules of regularly buying & selling some of the securities.

OTOH, he agrees that index investing has the advantage of reduced fees. I guess an overall conclusion would be that a low-fee, buy&hold fund of some kind would be the best thing for small investors.

grantmeaname

  • Magnum Stache
  • ******
  • Posts: 4811
  • Age: 27
  • Location: NYC
  • Cast me away from yesterday's things
Re: Efficient market hypothesis, index investing and value investing
« Reply #36 on: February 03, 2013, 11:48:40 AM »
I think he's overstating his case to get pageviews.

He's saying that some indices are not passively managed. ("Many indices are relatively actively managed" is about as noncommittal as a statement can get while still saying anything.) I think that the number of companies that move off of the S&P500 are rarely significant in a given year, and even more insignificant for an index like the Russell 3000. Considering that the last 25 or 50 companies of the Russell 3000 are like a tenth of a percent of the index's market cap, I don't think that's significant. And then there are indices like the MSCI US Broad Market index, which doesn't really have stocks moving off of it; Vanguard's VTSMX and VTSAX are based on it and are the firm's most popular funds by a wide margin, so I don't think it should be discounted with a handwaving "Many ... are relatively" and a shrug.

davidm

  • 5 O'Clock Shadow
  • *
  • Posts: 5
Re: Efficient market hypothesis, index investing and value investing
« Reply #37 on: February 03, 2013, 02:44:20 PM »
I think the argument between passive and active managment is one which clearly cannot be answered simply yes or no. Rather like saying momentum investing is better or worse than value investing, mean-reversion models, contrarian signals etc etc. It depends on the market or asset class you are considering, your timeframe and what liability you are trying to hedge.

Active management can clearly generate superior returns in selected instances. Take the concrete example of the Yale endowment. Over the past 20 years, Yale has used almost pure active management strategies to outperform other endowments by an average of 5.2%/annum, generating a real return (above CPI) of around 7.3%. Of that 5.2%, 1.1% (20%) is attributable to asset allocation and 4.1% (80%) to the choice of active manager. Note, however, that Yales allocation to US and global stocks is only around 14% and to fixed income only 4%. The bulk of their allocation is to hedge funds and private equity for the reason that they believe that manager selection can still exploit market inefficiencies in these areas whilst such an approach is far harder to achieve with long-only equity or bond managers. The problem for a retail investor is that they simply don't have the quality of information that Yale has regarding active managers or their ability to negotiate discounts on fee structures.

I remember reading in 1998 (yes long time ago!) a research paper (from JPMorgan) that looked at the information ratios of long-only active managers for various asset classes. It found that active managers indexed vs. US equity indices had average information ratios (IR) of almost exactly zero (pre fees). US corporate bond managers had average IR of 0.3. However, the average active managers indexed vs. global government bond indices had very solid IRs of around 0.5. The argument for the difference was that a) global bond markets were (at that time) less efficient than the US stock markets b) that the disinflationary impetus from globalization combined with the rise of central bank inflation targetting had created a clear opportunity for active managers to maintain betas > 1 with indices with a high degree of conviction. As a result the conclusion was that there was little point choosing active managers of US stocks but it was very worthwhile attempting to select actively managed government bond funds. Forward 15 years and whilst I haven't seen an equivalent paper, I would posit that it would be far harder to argue for selecting a actively managed bond fund with global bond yields at all time lows, cross-market correlations at very high levels and the efficiency of G10 bond markets far higher. It would be far easier, however, to justify choosing active management for an emerging market bond fund where there are far more structural breaks and market inefficiencies to exploit.



grantmeaname

  • Magnum Stache
  • ******
  • Posts: 4811
  • Age: 27
  • Location: NYC
  • Cast me away from yesterday's things
Re: Efficient market hypothesis, index investing and value investing
« Reply #38 on: February 03, 2013, 07:01:21 PM »
Active management can clearly generate superior returns in selected instances. Take the concrete example of the Yale endowment. Over the past 20 years, Yale has used almost pure active management strategies to outperform other endowments by an average of 5.2%/annum, generating a real return (above CPI) of around 7.3%. Of that 5.2%, 1.1% (20%) is attributable to asset allocation and 4.1% (80%) to the choice of active manager. Note, however, that Yales allocation to US and global stocks is only around 14% and to fixed income only 4%. The bulk of their allocation is to hedge funds and private equity for the reason that they believe that manager selection can still exploit market inefficiencies in these areas whilst such an approach is far harder to achieve with long-only equity or bond managers. The problem for a retail investor is that they simply don't have the quality of information that Yale has regarding active managers or their ability to negotiate discounts on fee structures.
That seems like a pretty good summary of the section I read of Pioneering Portfolio Management.

Nords

  • Magnum Stache
  • ******
  • Posts: 3198
  • Age: 59
  • Location: Oahu
    • Military Retirement & Financial Independence blog
Re: Efficient market hypothesis, index investing and value investing
« Reply #39 on: February 03, 2013, 08:08:06 PM »
Active management can clearly generate superior returns in selected instances. Take the concrete example of the Yale endowment. Over the past 20 years, Yale has used almost pure active management strategies to outperform other endowments by an average of 5.2%/annum, generating a real return (above CPI) of around 7.3%. Of that 5.2%, 1.1% (20%) is attributable to asset allocation and 4.1% (80%) to the choice of active manager.
Warren Buffett would be another example, or Tweedy-Browne. 

Fraud & insider trading are also hallmarks of exceptional portfolio performance.  Have we figured out yet which category Yale's success falls into, compared to how long other "stellar investors" managed to rack up their jail sentences records?