A Random Walk Down Wall Street

I began reading A Random Walk Down Wall Street a few days ago and am wondering, if I can gain anything out of it. I find it to be a very interesting read, but it seems like everything about random walk theory is consistent with the efficient market hypothesis and regardless of being right or wrong, I don't really see a situation in which a strong believer in EMH is considered a valuable asset on Wall Street. Is it worth the read simply for a history lesson, even if I need to forget about random walk theory if I ever want to work for a hedge fund?

 

You need to listen to W, Read X, Disregard Y, And come to the conclusion Z. (excuse my sarcasm)

But seriously I think the only one still trying to defend it is Fama himself and even him- if he's poked enough- will say things like ""poorly informed investors could theoretically lead the market astray" and that stock prices could become irrational"

It's like reading the theory of medicinal leeches to prepare for medical school in the 21st century. on the other hand, reading good books and thinking critically for yourself- that's invaluable imo.

 
<span class=keyword_link><a href=/resources/skills/economics/seigniorage target=_blank>Seigniorage</a></span>:
But seriously I think the only one still trying to defend it is Fama himself and even him- if he's poked enough- will say things like ""poorly informed investors could theoretically lead the market astray" and that stock prices could become irrational"

It's like reading the theory of medicinal leeches to prepare for medical school in the 21st century. on the other hand, reading good books and thinking critically for yourself- that's invaluable imo.

Irrational =/= inefficient And Fama is not the only one who still defends it. And comparing it to medicinal leeches is absurd. I don't think a doctor is constantly worrying about "How are medicinal leeches allowing such an opportunity/trade/injury/disease to arise?".
 

The EMH concept is clearly wrong in the most rigid sense that Malkiel has argued. In fact all of modern portfolio theory is wrong. You can't have people like Buffett if those concepts are true. I could elaborate, but basically you should read the Super Investors of Graham and Doddsville. To quote Buffett, some of these concepts are "Alice in Wonderland." Whatever people want to say, the fact is that there are some investors who are smarter, faster and just better, and who can consistently make out sized returns.

BUT that being said, the stock market is extremely competitive. Understanding that and having respect for it is a key lesson from A Random Walk. Generally, the larger the cap and the better followed the company, the more efficient that stock will be. Smaller cap stocks are much less efficient but also much more dangerous -- you get worse management, wider swings, less transparency, more business risk.

 
Ravenous:
The EMH concept is clearly wrong in the most rigid sense that Malkiel has argued. In fact all of modern portfolio theory is wrong. You can't have people like Buffett if those concepts are true. I could elaborate, but basically you should read the Super Investors of Graham and Doddsville. To quote Buffett, some of these concepts are "Alice in Wonderland." Whatever people want to say, the fact is that there are some investors who are smarter, faster and just better, and who can consistently make out sized returns.

BUT that being said, the stock market is extremely competitive. Understanding that and having respect for it is a key lesson from A Random Walk. Generally, the larger the cap and the better followed the company, the more efficient that stock will be. Smaller cap stocks are much less efficient but also much more dangerous -- you get worse management, wider swings, less transparency, more business risk.

From my understanding, the EMH'ers out there would say Buffett is just part of the distribution curve of investors, and that you can't attribute his performance to luck vs skill. I think that's horseshit, but opinions are opinions.

OP, read the book so you know what the arguments are. You should be conversant in all the major topics, at the very least so you can refute them, or even better, trade against them.

 
Best Response

The market is not entirely random (you can find anomalies, especially in hindsight) but it's random enough that it is almost impossible to extract money from it on a consistent basis.

Traders that make consistent $ tend to take massive tail-risks (martingale strategies, positive carry trades, writing out of the money options, etc.)... make $ under normal conditions for a few years, and then give it all back + some in the matter of weeks. I believe Taleb said that no bank in history has ever made money in the long-term... there's a point where it blow-ups and loses more money than it had ever made.

The opportunities where traders exploit legitimate anomalies (Ed Thorp - convertible arbitrage) are very short-lived. Crowding is inevitable. LTCM had a great strategy but when Wall Street recognized the game LTCM had to become massively leveraged to achieve alpha... and they blew up because of liquidity issues/over-leverage and not because of a bad strategy.

 
mb666:
LTCM had a great strategy but when Wall Street recognized the game LTCM had to become massively leveraged to achieve alpha... and they blew up because of liquidity issues/over-leverage and not because of a bad strategy.
This is simply false. They did not massively increase leverage to boost alpha.
 
Dr Joe:
mb666:
LTCM had a great strategy but when Wall Street recognized the game LTCM had to become massively leveraged to achieve alpha... and they blew up because of liquidity issues/over-leverage and not because of a bad strategy.
This is simply false. They did not massively increase leverage to boost alpha.

They eventually became leveraged 100:1 and weren't charged a haircut. That's not how they operated during their initial launch. 100:1 for a portfolio of convergence trades, regardless whether in different markets, is absurd. It's in Lowenstein's book.

 

LTCM did not have a great strategy that was recognized and dissapeared because the market is efficient...they just sold a bunch of options, adding no value, and eventually when volatility picked up they got blown out. We all could create the same PnL profile by levering up and selling out of the money puts on the S&P and making money until the next bust.

The market is massively inefficient....even if one believes that a free market would be efficient (which i dont), we have so many non-profit-maximizing actors in markets that it isnt even close to free and unfetered. Let's start with the fact that short term interest rates are set by twelve guys who meet in a room every six weeks....

 
Bondarb:
LTCM did not have a great strategy that was recognized and dissapeared because the market is efficient...they just sold a bunch of options, adding no value, and eventually when volatility picked up they got blown out. We all could create the same PnL profile by levering up and selling out of the money puts on the S&P and making money until the next bust.

The market is massively inefficient....even if one believes that a free market would be efficient (which i dont), we have so many non-profit-maximizing actors in markets that it isnt even close to free and unfetered. Let's start with the fact that short term interest rates are set by twelve guys who meet in a room every six weeks....

  1. How can you seriously claim that LTCM was just writing options? The partners, perhaps better than anyone else in the world, would have known this to be a sure way to get wiped out, and would not have had the vast majority of their own capital invested in the fund if this were the case. While I agree that a rational agent might pursue such a strategy if he can extract 2% + 20% out of it from the investors, he would certainly not have all (or much) of his money invested in it. They were doing much more than writing options.

  2. The presence of non-profit-maximizing actors does not in of itself make a market inefficient. I assume that by profit-maximizing you meant rational, since technically none of us seek to purely maximize profits. But even the presence of irrational investors does not prevent a market from being efficient. In fact, irrational investors are almost a necessary condition for a market to exist. For example, they are vital to the theoretical derivation of a bid-ask spread, amongst other things.

  3. How does the fact that short-term rates are set by a 12 guys in a room make the market inefficient?

Ravenous:
The EMH concept is clearly wrong in the most rigid sense that Malkiel has argued. In fact all of modern portfolio theory is wrong. You can't have people like Buffett if those concepts are true. I could elaborate, but basically you should read the Super Investors of Graham and Doddsville. To quote Buffett, some of these concepts are "Alice in Wonderland." Whatever people want to say, the fact is that there are some investors who are smarter, faster and just better, and who can consistently make out sized returns.
  1. Ravenous - sure you can have people like Buffet while allowing for EMH. He can be a coin-flipping orangutan, and don't tell me about how long his track record is. You should think about the independence of returns on his investments. Sample size decreases astronomically once you allow for intertemporal dependence and he is no longer the 10-sigma outlier that he appears to be. Pompous orangutan.
 
Bondarb:
LTCM did not have a great strategy that was recognized and dissapeared because the market is efficient...they just sold a bunch of options, adding no value, and eventually when volatility picked up they got blown out. We all could create the same PnL profile by levering up and selling out of the money puts on the S&P and making money until the next bust.

The market is massively inefficient....even if one believes that a free market would be efficient (which i dont), we have so many non-profit-maximizing actors in markets that it isnt even close to free and unfetered. Let's start with the fact that short term interest rates are set by twelve guys who meet in a room every six weeks....

From my understanding wasn't LTCM also exploiting arbitrage opportunities in the world bond markets? I thought they were buying up "old" debt in favor of newly issued as they viewed it as under priced?

"Life all comes down to a few moments. This is one of them." - Bud Fox
 
Bondarb:
LTCM did not have a great strategy that was recognized and dissapeared because the market is efficient...they just sold a bunch of options, adding no value, and eventually when volatility picked up they got blown out. We all could create the same PnL profile by levering up and selling out of the money puts on the S&P and making money until the next bust.

The market is massively inefficient....even if one believes that a free market would be efficient (which i dont), we have so many non-profit-maximizing actors in markets that it isnt even close to free and unfetered. Let's start with the fact that short term interest rates are set by twelve guys who meet in a room every six weeks....

LTCM's initial strategy was essentially a risk-free arbitrage operation where they bought an old-issue bond and shorted a newly issued bond (of same maturity)... held to maturity they would capture the spread. When they started the spread was wide and the strategy was very profitable with limited risk. However, Wall Street eventually discovered this strategy and the trade became crowded. LTCM foolishly became way over-leveraged to get alpha from greatly diminished spreads. Prime-brokers didn't even charge a haircut so the leverage was insane. Because of increased competition LTCM also ventured into less liquid markets (e.g. Russia), but all in all they were market-neutral.

LTCM failed because of liquidity. If they held their positions to maturity (after Russia's default) they actually would've made money.

 
mb666:
Bondarb:
LTCM did not have a great strategy that was recognized and dissapeared because the market is efficient...they just sold a bunch of options, adding no value, and eventually when volatility picked up they got blown out. We all could create the same PnL profile by levering up and selling out of the money puts on the S&P and making money until the next bust.

The market is massively inefficient....even if one believes that a free market would be efficient (which i dont), we have so many non-profit-maximizing actors in markets that it isnt even close to free and unfetered. Let's start with the fact that short term interest rates are set by twelve guys who meet in a room every six weeks....

LTCM's initial strategy was essentially a risk-free arbitrage operation where they bought an old-issue bond and shorted a newly issued bond (of same maturity)... held to maturity they would capture the spread. When they started the spread was wide and the strategy was very profitable with limited risk. However, Wall Street eventually discovered this strategy and the trade became crowded. LTCM foolishly became way over-leveraged to get alpha from greatly diminished spreads. Prime-brokers didn't even charge a haircut so the leverage was insane. Because of increased competition LTCM also ventured into less liquid markets (e.g. Russia), but all in all they were market-neutral.

LTCM failed because of liquidity. If they held their positions to maturity (after Russia's default) they actually would've made money.

The trade u describe is not even close to a risk-free arbitrage. It is just a fixed income relative value trade and I have done many of them in my career and used to work for a fund that did nothing but that type of trading. I am not going to go into why this trade isnt riskless, especially when levered up enough to "move the needle" at a large hedge fund, but it contains a ton of risk and it isnt some genius idea that no one had thought of before.

 
mb666:
LTCM failed because of liquidity. If they held their positions to maturity (after Russia's default) they actually would've made money.
Not entirely true - their positions did not end up being all that great - hardly breakeven after a few years. And Russia's default had much less to do with it than popularly thought.
 

Perhaps I wasnt being clear...LTCM wasnt actually writing options, but their positions mimicked a blind option selling strategy. They did a bunch of fixed income relative value trades that relied on reversion to the mean and low volatility, levered up and got way too big, and then when vol spiked they got wiped out. And BTW the principals at that firm didnt know anything "more then anybody" about trading...they were unproven traders who used their academic backgrounds as a marketing gimick to raise money.

 
Bondarb:
Perhaps I wasnt being clear...LTCM wasnt actually writing options, but their positions mimicked a blind option selling strategy. They did a bunch of fixed income relative value trades that relied on reversion to the mean and low volatility, levered up and got way too big, and then when vol spiked they got wiped out. And BTW the principals at that firm didnt know anything "more then anybody" about trading...they were unproven traders who used their academic backgrounds as a marketing gimick to raise money.
You are right - Meriwether was an unproven trader that used academic background as a marketing gimmick.

I guess I forgot about Salomon's PhD program.

Edit - Bondarb - don't take my tone to be directed personally at you - I know you are one of the best contributors to the site. But I disagree pretty strongly with your views here; I am not sure you facts are 100% correct.

 

Buffett isn't flipping coins though. There is a system backed up by a common sense approach. The 10 sigma argument doesn't stand up to even basic logical thinking. The fact that the system can be reproduced by others invalidates the claim that Buffett was merely lucky. Greenblatt reproduced it to the tune of 40% compound annual returns, as have more than a dozen others in the 20 to 40% range over multi-decade time periods.

The funny thing about LTCM is that they knew they were effectively selling options -- "picking up nickels in front of bulldozers" -- and they did it anyway. The math said a severe black swan event was very unlikely so they took the chance. Unfortunately for purely mathematical approaches, those kind of extreme tail events are happening more and more frequently, a lesson that John Meriwether seems to enjoy re-learning (apparently).

 

Thanks for the help guys, it looks like I will continue with my book, but don't really have much of an opinion on the ensuing argument, because I still view my opinion on the topic as developing and indecisive, so I don't want to say something that I would disagree with tomorrow, however:

Ravenous:
Greenblatt reproduced it to the tune of 40% compound annual returns, as have more than a dozen others in the 20 to 40% range over multi-decade time periods.

I admit, I did get this information from Wikipedia, but if Greenblatt really did receive most of his seed money from Michael Milken, that raises the question if he was a value investing guru or part of something much more shady.

Competition is a sin. -John D. Rockefeller
 
Ravenous:
Buffett isn't flipping coins though. There is a system backed up by a common sense approach. The 10 sigma argument doesn't stand up to even basic logical thinking. The fact that the system can be reproduced by others invalidates the claim that Buffett was merely lucky. Greenblatt reproduced it to the tune of 40% compound annual returns, as have more than a dozen others in the 20 to 40% range over multi-decade time periods.

The funny thing about LTCM is that they knew they were effectively selling options -- "picking up nickels in front of bulldozers" -- and they did it anyway. The math said a severe black swan event was very unlikely so they took the chance. Unfortunately for purely mathematical approaches, those kind of extreme tail events are happening more and more frequently, a lesson that John Meriwether seems to enjoy re-learning (apparently).

My point about returns being inter-temporally correlated - what would cause this? Using the same strategy through time. So he is lucky that his strategy worked. Think of it this way: rather than flipping a coin to determine your investment returns every day/month/year, you initially pick one out of, lets say, 1000 possible strategies. 30 or so years later, it turns out that your strategy was (perhaps) the best of the 1000, and you claim that your chance of such success was not 1 in 1000, but rather 1 in 2^(12*30) or something similarly absurd. Well, the point is, you just picked (randomly or not) the strategy that ended up performing very well, and you declare yourself an investing genius ex-post, but ex-ante any of the other strategies could have performed just as well.

In such a framework, the fact that others have been successful with it should not be surprising - if you invest in a strategy at some point, and at some later point it turns out to have worked well, it is obvious that all those that picked that strategy will have done well.

With that in mind, can you really make an argument for Buffett generating alpha in the future (which would imply inefficiencies in the market), while granting that past returns do not predict future returns?

Furthermore, simple no-quasi-arbitrage arguments indicate that if Buffett really was expected to make excess returns, others replicating his strategy would drive prices of assets he likes through the roof, prohibiting him from finding attractive investment opportunities.

Lastly, regarding you LTCM comment, I suggest taking Taleb with a grain of salt. First of all, the fall of LTCM really was not such an extreme tail event - similar (and greater) market turbulence had been observed in the 10 years leading up to the crash. Secondly, they were doing much more than selling options - they were dynamically replicating them and taking the spread between the actual price and the (lower) cost of replication. But this wasn't even the main thing that got LTCM in trouble. The "picking up nickels in front of bulldozers" comment makes for a nice catchy phrase but is in reality quite false. And lastly, purely mathematical approaches are very well aware of extreme tail events, but of course the problem with calibrating such models is the limited history available, as well as the possibilities of events so drastic they have never happened.

 

My point about returns being inter-temporally correlated - what would cause this? Using the same strategy through time. So he is lucky that his strategy worked. Think of it this way: rather than flipping a coin to determine your investment returns every day/month/year, you initially pick one out of, lets say, 1000 possible strategies. 30 or so years later, it turns out that your strategy was (perhaps) the best of the 1000, and you claim that your chance of such success was not 1 in 1000, but rather 1 in 2^(12*30) or something similarly absurd. Well, the point is, you just picked (randomly or not) the strategy that ended up performing very well, and you declare yourself an investing genius ex-post, but ex-ante any of the other strategies could have performed just as well.


This is where your thinking is flawed. The other strategies could not have performed just as well. Your argument is that any of the 1,000 strategies could have worked. Stock values are connected to underlying businesses, which are subject to economic truths. Capitalism is a system with predictable consequences. It's so predictable in fact that mother nature (evolution) is based on the same set of constructs. Buffett relentlessly studied these concepts from the age of 7 to the exclusion of most other considerations, and then had the discipline to apply what he learned consistently over a multi-decade period. A lot of his money was made on a few simple bets. Arguing that the success of his investment in Coca Cola, for example, was luck is absurd.

My point is, it's not the same strategy in the static, purely mathematical sense that you are suggesting. It's a doctrine. A body of knowledge, or an understanding of the way the world works, and what that means for individual stock returns. And Buffett is a master at that. No person could apply that "strategy" in a static, clinical way and expect the same sort of results. In other words, the mathematical models fail (and EMH is wrong) because the world cannot be reduced to a formula. But you can count on basic principles, which is what Buffett does.


In such a framework, the fact that others have been successful with it should not be surprising - if you invest in a strategy at some point, and at some later point it turns out to have worked well, it is obvious that all those that picked that strategy will have done well.

But they didn't pick the same strategy, they all used variations on the same themes, which is a totally different argument.


Lastly, regarding you LTCM comment, I suggest taking Taleb with a grain of salt. First of all, the fall of LTCM really was not such an extreme tail event - similar (and greater) market turbulence had been observed in the 10 years leading up to the crash. Secondly, they were doing much more than selling options - they were dynamically replicating them and taking the spread between the actual price and the (lower) cost of replication. But this wasn't even the main thing that got LTCM in trouble. The "picking up nickels in front of bulldozers" comment makes for a nice catchy phrase but is in reality quite false. And lastly, purely mathematical approaches are very well aware of extreme tail events, but of course the problem with calibrating such models is the limited history available, as well as the possibilities of events so drastic they have never happened.

I haven't read Taleb. What got them in trouble was placing bets with extreme leverage and no way to quickly unwind those bets. They went short volatility and lost (predictably). It's a great strategy until it doesn't work, and then it doesn't work at all. If you can't model for events in global markets that are not particularly uncommon over a long time period (sovereign default), then what good are the models?

 
Ravenous:
No, it doesn't raise that question. Try again.

Michael Milken was indicted on 98 counts of racketeering and securities fraud, so I'm supposed to assume a guy that received all their seed money from a guy like him is 100% legitimate and never traded on an insider tip?

Competition is a sin. -John D. Rockefeller
 
Hooked on LEAPS:
Ravenous:
No, it doesn't raise that question. Try again.

Michael Milken was indicted on 98 counts of racketeering and securities fraud, so I'm supposed to assume a guy that received all their seed money from a guy like him is 100% legitimate and never traded on an insider tip?

He has a winning, audited track record dating to 1985. Don't be an asshole.

 

I think you should continue to read the book. I think there is an MIT professor who has written a more academically oriented book called a "non-random walk down wall street", but I haven't read that one yet...

The the concept of and discussion around EMH is an important to understand whether you agree with the hypothesis or not... you'll find many people who discuss the concept (for and against) have a surprisingly superficial understanding of it.

As you can see from this thread, it's very important know what hidden assumptions we make when discussing the performance of certain investors and strategies...

 

also, the statement "if they had been held to maturity the positions would have made money" does not equal "a riskless arbitrage"...risks surrounding financing of positions, liquidity, margins, haircuts, etc are very real. These arent risks that nobody had ever considered or heard of prior to LTCM and the trades they were doing were not some revelation that nobody else could see. If their trades were held to maturity they would have been profitable....and if a bull had tits it would be a cow.

Right now you can have a similar "riskless arbitrage"...just buy 3 year Italian debt with money you borrow from the ECB at 1% and lever massively. You will have big returns unless a) the mark to market losses are big enough that your prime broker demands more margin then you have b) your investors pull their money forcing you to liquidate c) Italy defaults. I think it is self-evident that this trade is not "risk free" but its the type of trade that will likely mature profitably and yet still could easily put you out of business (see MF Global).

 

Reading it is worthwhile also reading "Alchemy of FInance". Whatever value you have on EMH, there is model's people out there using it so you always need to know and understand your opponent.

Lot's of what happens in the market at certain times occurs because people are using EMH, you need to understand and identify those situations.

 

You guys are giving LTCM way to much credit if you think they were "dynamically replicating options and taking the spread between the actual price and the price of replication". They were just running a fixed income relative value book and betting that traditional relationships between bonds that had gotten out of whack would mean revert. Many other were doing the same thing at the same time and many are still doing it today. The only difference is that they had a good story about being nobel prize winners that helped them raise alot fo money and the only one with trading experience at the firm was a wild risk-taker who has blown up twice more since LTCM went under.

 
Bondarb:
You guys are giving LTCM way to much credit if you think they were "dynamically replicating options and taking the spread between the actual price and the price of replication". They were just running a fixed income relative value book and betting that traditional relationships between bonds that had gotten out of whack would mean revert. Many other were doing the same thing at the same time and many are still doing it today. The only difference is that they had a good story about being nobel prize winners that helped them raise alot fo money and the only one with trading experience at the firm was a wild risk-taker who has blown up twice more since LTCM went under.
Perhaps you are not giving them enough credit. I have personally spoken to several of the LTCM partners, and worked closely with one of them, and dynamic option replication was certainly a part of their investment approach. That being said, I acknowledge that I may be a bit biased.

Edit - and for what it's worth, I believe Meriwether is now at 3 blow-ups after LTCM. But he was not the only one with trading experience - more than half the partners had some form of experience.

 

Did you see any examples of these option replication trades? The term "option replication" can mean just about anything including just making bets on mean reversion across the yield curve...I do think they were replicating options in that way and in fact getting very short vol by synthetically selling options. I just dont believe they had any edge in it and certainly no more edge then the many other traders who were doing it then, do it now, and have solid long-term track records. Difference being that those guys just call it "bond trading" and they dont couch their strategies in terms that make them sound more complicated then they are.

 
Bondarb:
Did you see any examples of these option replication trades? The term "option replication" can mean just about anything including just making bets on mean reversion across the yield curve...I do think they were replicating options in that way and in fact getting very short vol by synthetically selling options. I just dont believe they had any edge in it and certainly no more edge then the many other traders who were doing it then, do it now, and have solid long-term track records. Difference being that those guys just call it "bond trading" and they dont couch their strategies in terms that make them sound more complicated then they are.

In a sense you are both right, when LTCM blew up they had hundreds of trades, including all the ones you two said. That said, the two biggest ones, and the ones that accounted for around 60% of their losses were the short vol and short european bond spreads, which is why they got screwed with the Russian default. Also if I'm not mistaken their vol trade was in equity markets, in which they had no experience but assumed they could apply their perfect models nonetheless. None of their strategies was riskless or an arbitrage of any kind, they simply used relative value trades, like thousands of traders nowadays. And in the final years they actually did pure directional trades, which is as risky as you can get. The thing that differentiated them from the rest was that their models were better, so they got it right more often than most people, and they managed to get a great leverage and no haircuts because of their great PR with the academics and nobel prize winners. Oh, and Meriweather and Hilibrand created JWM Partners 1 year after LTCM and they went under after the '08 crisis, so they clearly can't accurately predict extreme events. They can't be that good if they blew it twice.

 

Ravenous,

Just because Buffett's Coca Cola investment was successful ex-post (after the fact) and he has a story / rationalisation around it doesn't mean that the story / investment thesis was right ex-ante (before the investment) or that it is replicable. Buffett had equally compelling investment rationales for his failed investments before they failed. If you haven't, I suggest you read Snowball (Buffett's biography) to learn about the social aspects of why he made some of his investments like his Coke investment or in newspapers. He wanted to be associated with the major board members / shareholders / Allen & Co, etc...

Left fat tail events can happen to Coke (i.e. low probability high loss scenarios). A few off the top of my head: the EU stopping it from being sold due to new research on one of the chemicals in the product (this destroys the company's brand / moat)... A new CEO could take the shareholders for a ride Tyco style or alternatively start expanding into other business through acquisition thus changing the economics of owning the stock (this has happened with Coke before, they bought Columbia Pictures in the 80s than sold it when they realised how volatile it was)... the point being, on a long enough timeline (100 years as you suggested) the likelihood that you will encounter such a scenario is higher because you are effectively making the same bet over and over again (i.e. company specific non-systemic risk).

In my opinion, one of the reasons that Buffett has been so successful isn't necessarily an exceptional investing prowess, rather that he has an institutional set up (permanent capital & insurance float) that allows him to hold positions for a long time / ride out his bad investment decisions (contrast that with the LTCM situation).

If he was still running a hedge fund the way he did before berkshire he would have faced outflows in bad years and may have had to liquidate positions at inopportune times/losses. When Buffett makes a lousy investment he can hold it for a long time and milk the cash flows to reinvest in other businesses at a higher rate rather than having to close out positions at a loss to meet redemptions (i.e. the rate of return on cash flow from bad deals is higher than the bad deal IRR on a stand alone basis, thus increasing the returns on bad investments over time. His MIRR is better than his IRR on bad deals and he has the time/liquidity to benefit from this). He basically gets a second chance as long as his investments have decent cash flow and he still has a "Guru / Mystic / Oracle [Insert mythical / religious figure of choice] " status among his shareholders.

The useful part about discussing EMH, value investing, investment track records is when we consider the assumptions that we are implicitly making in our models of the world and what happens to the models predictions when we change those assumptions to reflect reality.

 

Relinquis,

It was right ex-ante in the sense that it had a very low probability of not working out and a very high probability of working with huge upside for clearly definable reasons. It's pretty extreme to try to reduce the Coke business model to a mere "story" that Buffett has made up like some kind of snake oil salesman. No equity investment is perfectly risk free if only because of the Tyco factor you mentioned. That doesn't prove anything about market efficiency though. He made a very high probability bet that paid off in multiples -- how can that be considered efficient? The very idea violates several principles of modern portfolio theory.

The institutional factors you mentioned are important. But it's hard to deny that studying investments for decades hasn't played a role in his success, even if that effort can't be directly quantified. I'm not sure why people always want to try to play down his work ethic. If anyone were to spend that much time on one complex activity free of institutional constraints that create inefficiencies (mutual fund rules, etc.), of course they would produce superior results.

I understand your argument about Buffett's permanent capital (and agree), but don't see how that is supposed to invalidate his success as he built the model from scratch and was smart enough to come up with it in the first place. To the extent that he found a way to consistently beat the market legally, doesn't that pour cold water on EMH? Because EMH says that no market participant can legally obtain information that will lead to consistent out performance, but then detractors try to add qualifiers to that statement -- "Yes, but Buffett is doing it because he has a longer time horizon." No qualifiers necessary. What Buffett's model really shows is that markets are not only not perfectly efficient (although they are very competitive in some places), but that a large part of the mirage of efficiency is probably due to artificial constraints on the investing process (the obsession with quarterly returns, narrow focus on particular stocks or industries or capitalization sizes, etc.) that push investors into the most competitive parts of the market.

So I disagree with you or anyone who supports EMH in the absolute sense. The only value I see in reading Malkiel's book or being aware of EMH is that one should be aware how competitive the market is. There are ways to generate consistent out performance over time but they are not easy to achieve and are not likely accessible to the average investor at home with an E-Trade account.

 
Ravenous:
Relinquis,

It was right ex-ante in the sense that it had a very low probability of not working out and a very high probability of working with huge upside for clearly definable reasons.

You are estimating the probabilities and up/downsides with knowledge of the outcome. Without such knowledge (ex-ante) if it was such an obvious bet then others would have piled on, driving price up and returns down...
Ravenous:
It's pretty extreme to try to reduce the Coke business model to a mere "story" that Buffett has made up like some kind of snake oil salesman. No equity investment is perfectly risk free if only because of the Tyco factor you mentioned. That doesn't prove anything about market efficiency though. He made a very high probability bet that paid off in multiples -- how can that be considered efficient? The very idea violates several principles of modern portfolio theory.
The fact that a certain bet happened to pay off violates... what exactly??
Ravenous:
The institutional factors you mentioned are important. But it's hard to deny that studying investments for decades hasn't played a role in his success, even if that effort can't be directly quantified. I'm not sure why people always want to try to play down his work ethic. If anyone were to spend that much time on one complex activity free of institutional constraints that create inefficiencies (mutual fund rules, etc.), of course they would produce superior results.
Think about equity research analysts who study only a few stocks for many many years, and even they can't cut it as traders. The study of fundamentals does not create alpha (unless cost of studying exceeds achievable alpha, in which case you are at a loss anyway).
Ravenous:
I understand your argument about Buffett's permanent capital (and agree), but don't see how that is supposed to invalidate his success as he built the model from scratch and was smart enough to come up with it in the first place. To the extent that he found a way to consistently beat the market legally, doesn't that pour cold water on EMH? Because EMH says that no market participant can legally obtain information that will lead to consistent out performance, but then detractors try to add qualifiers to that statement -- "Yes, but Buffett is doing it because he has a longer time horizon."
First of all, absolute returns don't matter - risk-adjusted returns matter. I agree that the time horizon qualifier is not necessary.
Ravenous:
No qualifiers necessary. What Buffett's model really shows is that markets are not only not perfectly efficient (although they are very competitive in some places), but that a large part of the mirage of efficiency is probably due to artificial constraints on the investing process (the obsession with quarterly returns, narrow focus on particular stocks or industries or capitalization sizes, etc.) that push investors into the most competitive parts of the market.
If the above list of constraints caused inefficiencies, then they would be changed.
 
Dr Joe:
Ravenous:
Relinquis,

It was right ex-ante in the sense that it had a very low probability of not working out and a very high probability of working with huge upside for clearly definable reasons.

You are estimating the probabilities and up/downsides with knowledge of the outcome. Without such knowledge (ex-ante) if it was such an obvious bet then others would have piled on, driving price up and returns down...
Ravenous:
It's pretty extreme to try to reduce the Coke business model to a mere "story" that Buffett has made up like some kind of snake oil salesman. No equity investment is perfectly risk free if only because of the Tyco factor you mentioned. That doesn't prove anything about market efficiency though. He made a very high probability bet that paid off in multiples -- how can that be considered efficient? The very idea violates several principles of modern portfolio theory.
The fact that a certain bet happened to pay off violates... what exactly??
Ravenous:
The institutional factors you mentioned are important. But it's hard to deny that studying investments for decades hasn't played a role in his success, even if that effort can't be directly quantified. I'm not sure why people always want to try to play down his work ethic. If anyone were to spend that much time on one complex activity free of institutional constraints that create inefficiencies (mutual fund rules, etc.), of course they would produce superior results.
Think about equity research analysts who study only a few stocks for many many years, and even they can't cut it as traders. The study of fundamentals does not create alpha (unless cost of studying exceeds achievable alpha, in which case you are at a loss anyway).
Ravenous:
I understand your argument about Buffett's permanent capital (and agree), but don't see how that is supposed to invalidate his success as he built the model from scratch and was smart enough to come up with it in the first place. To the extent that he found a way to consistently beat the market legally, doesn't that pour cold water on EMH? Because EMH says that no market participant can legally obtain information that will lead to consistent out performance, but then detractors try to add qualifiers to that statement -- "Yes, but Buffett is doing it because he has a longer time horizon."
First of all, absolute returns don't matter - risk-adjusted returns matter. I agree that the time horizon qualifier is not necessary.
Ravenous:
No qualifiers necessary. What Buffett's model really shows is that markets are not only not perfectly efficient (although they are very competitive in some places), but that a large part of the mirage of efficiency is probably due to artificial constraints on the investing process (the obsession with quarterly returns, narrow focus on particular stocks or industries or capitalization sizes, etc.) that push investors into the most competitive parts of the market.
If the above list of constraints caused inefficiencies, then they would be changed.

Your argument is circular. You say that if it was so obvious, others would have acted. It was obvious, even without ex-post knowledge, and they did not act. Who knows why, but they didn't. There are literally thousands of other cases that could be cited where an outcome was obvious and only a few acted. That's what investing is.

The probability of Coke stock going to zero is extremely low. You give examples of some strange, extreme tail scenarios that are so improbable that they are effectively at zero probability. The US is not going to outlaw Coke, just like it has not outlawed cigarettes, alcohol, automobiles, airline travel, household cleaning products, and electricity. If Coke is a depressed valuation due to high interest rates and the bad mood of the general investing public, and it's not going to zero, and you know with near-certainty that it is going to grow indefinitely, that is a great investment, and it is an example of how the market is not perfectly efficient.

It violates MPT because MPT says that risk and return are always correlated, which is clearly not the case. It's one of the reasons the market cannot be perfectly efficient. If you have a claim as to why EMH holds true 100% of the time (which is what we are discussing), then feel free to enlighten everyone else. At this point, it seems that you do not have a claim.

It's clear that you know little about fundamental analysis. Equity research analysts follow 5 or at most 10 stocks. What if those stocks are well priced over a decade long period? Those analysts will have nothing to do. That scenario is completely different from Buffett (or any other well schooled generalist investor with long tenure) who is evaluating stocks across all industries, geographies and cap sizes. There are 10,000 stocks alone in the US, and multiples of that world wide.

You claim that no one can do that (fundamental research doesn't produce alpha) because you haven't seen it done. There are dozens of proven track records that speak otherwise. Buffett did it, Greenblatt did it, and I work for someone who has done it. Interestingly, all three systems are pretty similar. I am not sure why other people don't learn (the approaches / strategies have been made public) and apply what works, but apparently they don't. I suspect they don't because they are impatient and/or not dedicated enough to make it work (it probably takes a decade of intense study before you would be able to succeed).

Yes, risk adjusted returns matter. That's the point I made above with MPT.

You would think that the constraints would be changed, but they have not been. Mutual funds still focus on quarterly returns. They still limit themselves based on cap size and minimum dollar value of the shares (no buying under $5). There are others. These restrictions do not serve the mutual fund shareholders well, but they still exist, and they have a noticeable impact on the market.

In a perfect world, everything you are saying would be true. But the market doesn't work that way, and that's where the opportunities exist. If you had the opportunity to work with a skilled investor, you would be able to see that first hand.

At any rate, we don't agree. EMH does not hold in real life, and I'm not even sure it holds in academia without some bizarre assumptions.

 

I'm very glad that people on WSO realize that EMH is useful from an academic perspective, but not practical in real life.

You need to know about EMH just as being part of an educated person in finance. That said, the only time when being a diehard orthodox believer in EMH would matter is if you're working at Vanguard, but that's it.

 

I'm not going to convince you Dr. Joe, but for the sake of discussion I will respond to a few of your points.

I'm not up on Coke numbers so I wasn't claiming it was a phenomenal value today, though it might be. But it was in 1988 when Buffett bought it. The reason it was a phenomenal value is that it had very little downside with large upside as a growing annuity -- I don't know what the valuation was, but presumably it was on the lower end of its historic P/E range, which is why he was buying it. I don't have access to whatever sell side analysts were writing about Coke in 1988 but I can guess with a high degree of confidence that it was focused on next quarter's and next year's earnings estimates, or maybe whatever the then-CEO said at the latest investor conference (or whatever). Almost assuredly, no one sat down and looked at what kind of company Coke might be 10, 20 or 30 years down the line. The sell side doesn't look that far in advance, and to be fair, most other investors don't either.

What made Coke such a great investment is that it could grow for decades to come with no crazy assumptions or hail mary business plans. In existing markets, it would grow at the rate of inflation, population growth, +/- market share gains/losses (though certainly gains are more likely over time). That may only be a few % per year, but it is a reliable few % and it compounds well over long time periods. And that doesn't include developing markets where the growth is much higher. What makes it such a great business is that it doesn't require a lot of capital, but produces a lot of cash flow -- which can be returned to shareholders, or reinvested at high compounding rates. Once the company has its distribution built out, it can't be unseated, which means the cash flows are an annuity that deserve a low discount rate.

I would have to think about a good list, but it could be one of the 10 best businesses in the world, almost certainly in the top 25. In 1988, Buffet was 58 years old. He had looked at thousands of companies around the world. He would have known that Coke was one of the best business, and he would have known in less than an hour what the company's potential was (in fact, he had probably been waiting for some time to buy it). I'm actually not a huge Buffett fan boy, but having worked in the industry, I know with certainty that over such a long time period (50 years of effort) it would be more than possible to build up the skill required to make that kind of assessment with ease. Saying that it was only obvious that the investment would work out ex-post is an incredibly narrow, purely mathematical approach to the world -- flawed. This was no mere coin flip.

So for whatever reason, this supposedly perfectly efficient market did not discount the full potential of Coke's growth. I would agree that the sell side probably had "studied it so carefully" as you said above on a short-term basis, and it may have been perfectly efficient on that basis. But clearly it was not on a long-term basis. And since EMH says that all information that a rational investor would want to know is already impacted into stock prices (as you also alluded to above with your growth comments), then this should hold true in all cases. Clearly, it didn't. This is an easy example, but far from an isolated case.

I think it was "obvious" because the analysis on Coke was so rudimentary that you could have explained it to a 2nd grader. But it was only obvious if you were asking the right questions and if you had a very clear understanding of how business and investing work, coupled with an appropriate time frame (I say appropriate in the long-term sense). So in that sense, I agree that if all analysts were saying something about a stock with regard to a SHORT DURATION OUTLOOK, it would disappear. But the price may not include something that is off the radar of what the analysts are looking at. And they can miss things in the short-run as well. And that doesn't even include stocks that have no analyst coverage, which make up something like 80% of the US market.

Over some time period, say a decade, there are bound to be hundreds of ideas that are "obvious" on a multi-year basis if evaluated through the correct lens / method / approach. Over a life time, it could very well be thousands as I said before. Is every one going to increase 20x? Of course not. But a 2 or 3 year double is definitely better than a sharp stick in the eye. Under EMH, that shouldn't ever happen, yet some people seem to be able to make it happen fairly regularly (what if you had long-term capital and a system that specifically focused on targeting such ideas?). Granted, it's a life style -- you can't just sit down and make it happen. You have to live the markets and look at hundreds of companies before you can develop a good sense for what to look for.

My premise was / is that such investments violate EMH. Some people are able to find investments that fit the above criteria on a consistent ex-ante basis by scouring hundreds or thousands of potential investments. It's not easy though and most people don't have the discipline, personality, or ability. It's like what Buffett has said many times, "I could tell people exactly how to do it, and they still couldn't do it."

 

This thread is testament to the damage the EMH has done to both academia and real world finance. Why does there even need to be an argument over whether the EMH holds true or whether the market is efficient? My opinion is that it doesn’t actually matter. The EMH cannot be disproven. It’s like a zombie that doesn’t get down no matter how many mouthfuls of buckshot you give it. There is no control sample that says “this is what would have happened had asset prices fully reflected all available information” hence we cannot actually measure the level of efficiency. You can argue as much as you want whether Buffett is lucky or skilled, but neither side will win because how are you going to prove the other is wrong? Statistically speaking there is a chance Buffett could flip a coin and land on heads thousands of times in a row. Then again there is a chance he can persistently identify mispriced assets by skill. Fama will never receive a Nobel because his theory is completely useless.

I think the important thing to remember here is that despite everything posited by MPT, financial academia is actually thriving. There are many other theories, especially in the field of behavioural finance, which more accurately describe what really happens in the market without the need for unrealistic assumptions such as “perfect information”.

 

Benoit Mandlebrot in "The Misbehavior of Markets" disproves EMH, well the random walk/Brownian motion, from the standpoint that price moves are not entirely independent of previous fluctuations. Prices, even when measured statistically and proven beyond chance, do sometimes act dependently. The "sometimes" point, however, makes it difficult to profit consistently from the anomalies (usually requires hindsight to run the tests too). Volatility in particular tends to cluster. Wide range fluctuation days tend to occur consecutively.

I'm not sure if there has been mention of the semi-weak, semi-strong and weak-form. Depending on your definition of EMH there's a lot of gray area.It's tough to argue against the weak-form because even with information asymmetry managers still fail to consistently outperform the market because the people accessing the "better" information tends to vary (well perhaps anyone aside Steve Cohen from SAC).

Regarding LTCM I choose a poor word, as in risk-free, to describe the old/new debt strategy that would always result in a nominal profit if held to maturity. I suppose that high leverage was necessary to take advantage of the tight spreads so they were exposed to liquidity risk. I still contend that liquidity was the main reason for LTCM's demise and that Russia played the most significant role in triggering volatility spikes which essentially drained liquidity in a lot of their positions. Then collateral issues forced liquidation. I know that their portfolio consisted more than just the new/old debt type of trades. Some of their trades had no guarantee of convergence and were based off historical correlations... fixed-income/derivative trades that mocked equity pair trades (long KO/short PEP). For example I believe they placed bets on varying volatility maturites and among sovereign debt.

The whole six sigma argument to Russia's default is flawed because LTCM used ~4 years of historical data. If they used two decades worth of data then LTCM would've seen that Russia/Soviet Union had defaulted on their debt and probably would've been more evaluated such as a 2 sigma event and not a fat tail-risk one.

I can't believe that John Meriwhether is on his third hedge fund after blowing up the first two. Who is seeding this guy? Lewis made him sound like a genius at Soloman who was pushed out because of a scandal that he barely was responsible for. Can someone confirm the Wiki claim that his newest hedge fund was only about to raise ~$30,000,000?

 
mb666:
... I can't believe that John Meriwhether is on his third hedge fund after blowing up the first two. Who is seeding this guy? Lewis made him sound like a genius at Soloman who was pushed out because of a scandal that he barely was responsible for. Can someone confirm the Wiki claim that his newest hedge fund was only about to raise ~$30,000,000?

It's hard to overestimate the importance of name recognition and relationships in this industry.

 
mb666:
Regarding LTCM I choose a poor word, as in risk-free, to describe the old/new debt strategy that would always result in a nominal profit if held to maturity. I suppose that high leverage was necessary to take advantage of the tight spreads so they were exposed to liquidity risk. I still contend that liquidity was the main reason for LTCM's demise and that Russia played the most significant role in triggering volatility spikes which essentially drained liquidity in a lot of their positions. Then collateral issues forced liquidation. I know that their portfolio consisted more than just the new/old debt type of trades. Some of their trades had no guarantee of convergence and were based off historical correlations... fixed-income/derivative trades that mocked equity pair trades (long KO/short PEP). For example I believe they placed bets on varying volatility maturites and among sovereign debt.

The whole six sigma argument to Russia's default is flawed because LTCM used ~4 years of historical data. If they used two decades worth of data then LTCM would've seen that Russia/Soviet Union had defaulted on their debt and probably would've been more evaluated such as a 2 sigma event and not a fat tail-risk one.

I can't believe that John Meriwhether is on his third hedge fund after blowing up the first two. Who is seeding this guy? Lewis made him sound like a genius at Soloman who was pushed out because of a scandal that he barely was responsible for. Can someone confirm the Wiki claim that his newest hedge fund was only about to raise ~$30,000,000?

LTCM had stopped doing the old/new debt trade long before they collapsed. Too many people were doing it, and even with their leverage there was just no way to make big bucks. I agree that Russia triggered vol, but that was just one of the trades, they were still hemorrhaging money in many other ones. I completely disagree with the notion that it was a liquidity issue, the banks who bailed LTCM out kept the positions, and they barely broke even after smth like 6 years. They simply had very bad trades. When you trade you have a timeline. It's like saying that 1929 was a liquidity issue because in 1999 the stock market was higher, people simply couldn't keep their trades on long enough. I know it's a ridiculous example, but you get the point. That's what trading is about, managing your risk, and these guys did a horrible job at it, they were overconfident, had their heads up their asses and simply failed completely at predicting market movements in their trades.

BTW, I'm not 100% sure, but I think Russia had never defaulted in any of it's Ruble denominated debt, which is what happened there. It is extremely rare, since they could have basically printed money to pay the debt. Even if that would have caused hyperinflation, it was (and pretty much is) the norm. The trouble is with bonds issued in some other currency. So when you say that they only used 4 years and Russia had defaulted before that, you mean in Ruble debt? Can you check that? I'm really curious.

 
t4s:
EMH is the same as saying forget about making money baking bread, if there was money in it everybody would be already doing it.
And when you go into the baking business in a place that no one is baking, wouldn't you want to know why there is no one there?

Sure, perhaps no one has thought of opening a bakery there, but imagine a bakery could be opened there by anyone in the world at the click of a button without getting out of his chair. In fact, there are people whose job it is to look for specific places to open bakeries around the world. Thousands of such people, who have extensive baking experience, phenomenal education in baking (Baker scholars lol), and giant research teams assigned to the baking markets worldwide.

Then do you want to know why there is no bakery before investing your own capital?

 
Dr Joe:
Sure, perhaps no one has thought of opening a bakery there, but imagine a bakery could be opened there by anyone in the world at the click of a button without getting out of his chair. In fact, there are people whose job it is to look for specific places to open bakeries around the world. Thousands of such people, who have extensive baking experience, phenomenal education in baking (Baker scholars lol), and giant research teams assigned to the baking markets worldwide.

Then do you want to know why there is no bakery before investing your own capital?

Just wanted to say that usually you would get payed if you get a job at a bakery and most bakeries make money baking bread. Same for professional investors, eventhough, since rewards can be much higher, the environment is a lot more competitive and therefore most investors lose money.

But nobody is killing it baking bread. So the bread baking market would be near efficient, or efficient if you count in blue collar wages.

 

Cupiditate id dolorem est officiis blanditiis. Rem nobis harum et optio incidunt fugit veritatis animi.

Quis veritatis deserunt dignissimos quo facilis et commodi cum. Non veniam quasi hic alias dolorem libero. Sed quas omnis enim ut vitae amet. Minima quidem omnis facere qui quia non sunt.

Doloremque atque sequi incidunt amet. Et deleniti mollitia non soluta perferendis quo cumque placeat. Molestiae voluptatem accusantium in sit laudantium temporibus eos. Corrupti dignissimos dignissimos itaque consequatur enim. Omnis veniam non sequi dolorem nihil id autem. Quia eius aliquam porro et magni non. Consequatur dolor labore quidem non.

 

Repellat beatae repellat repellat cupiditate quia unde a. Quo quasi nemo sit qui ratione excepturi. Quia repellat commodi nihil nihil et officiis. Aut et molestiae consequatur eaque labore. Blanditiis dolores et vero impedit iure neque. In ea soluta nobis est voluptatem aut omnis qui. Aut itaque occaecati velit recusandae id.

Unde quod sequi ad. Ut vitae ut ipsum ullam recusandae cum nisi. Quos facilis et officia numquam.

Assumenda earum culpa alias delectus assumenda ducimus. Consequatur corporis vel expedita dolorem. Rem voluptatem perspiciatis et consequatur ea id. Corrupti quod aperiam sit eum veniam ut consequatur.

Career Advancement Opportunities

March 2024 Investment Banking

  • Jefferies & Company 02 99.4%
  • Goldman Sachs 19 98.8%
  • Harris Williams & Co. (++) 98.3%
  • Lazard Freres 02 97.7%
  • JPMorgan Chase 03 97.1%

Overall Employee Satisfaction

March 2024 Investment Banking

  • Harris Williams & Co. 18 99.4%
  • JPMorgan Chase 10 98.8%
  • Lazard Freres 05 98.3%
  • Morgan Stanley 07 97.7%
  • William Blair 03 97.1%

Professional Growth Opportunities

March 2024 Investment Banking

  • Lazard Freres 01 99.4%
  • Jefferies & Company 02 98.8%
  • Goldman Sachs 17 98.3%
  • Moelis & Company 07 97.7%
  • JPMorgan Chase 05 97.1%

Total Avg Compensation

March 2024 Investment Banking

  • Director/MD (5) $648
  • Vice President (19) $385
  • Associates (86) $261
  • 3rd+ Year Analyst (13) $181
  • Intern/Summer Associate (33) $170
  • 2nd Year Analyst (66) $168
  • 1st Year Analyst (202) $159
  • Intern/Summer Analyst (144) $101
notes
16 IB Interviews Notes

“... there’s no excuse to not take advantage of the resources out there available to you. Best value for your $ are the...”

Leaderboard

1
redever's picture
redever
99.2
2
BankonBanking's picture
BankonBanking
99.0
3
Betsy Massar's picture
Betsy Massar
99.0
4
Secyh62's picture
Secyh62
99.0
5
kanon's picture
kanon
98.9
6
DrApeman's picture
DrApeman
98.9
7
dosk17's picture
dosk17
98.9
8
GameTheory's picture
GameTheory
98.9
9
CompBanker's picture
CompBanker
98.9
10
Jamoldo's picture
Jamoldo
98.8
success
From 10 rejections to 1 dream investment banking internship

“... I believe it was the single biggest reason why I ended up with an offer...”