Too Much Information
 

http://www.vcconfidential.com/2007/0...ch-inform.html

Some of you may have seen this from my shared feed. I thought it was pretty interesting and somewhat counter to the way a lot of people feel about investing. It specifically made me think about someone like Barron, that chooses to use tons of data.

One interesting note is that the author talks about not only investments in small, relatively inefficient markets (venture capital), but also uses Warren Buffett as an example.

Perhaps Homer Simpson was right, "Facts are meaningless. You could use facts to prove anything that's even remotely true!"

Re: Too Much Information
 

Interesting read.

Re: Too Much Information
 

This is so true.

I concentrate on technical analysis which uses a derivative of one piece of data: price

Price is the current consensus of value amongst all market participants. It is the only really reliable piece of data available. It can't be massaged or manipulated like most other forms of data. (it can be manipulated to some degree, but this action can also be apparent in price action).

Markets tend to oscillate between greed and fear. These oscillations can be seen on hourly,daily, weekly, monthly charts.

Many TA also use too much information.They feel that the more indicators they use, the more accurate their trades will be. This is not usually the case however, as the more signals you rely upon, the greater the distance your analysis moves away from the pure data (price).

Re: Too Much Information
 

I'll give an example of one of my investments, I'm going to try to keep it as short as I can so just assume the things I say are true, you can look them up if you want.

Cavalier Homes(CAV) makes manufactured homes(MH, or also known as trailer houses). Since the 60's the MH industry has followed a predictable pattern relative to total housing starts in the US; whatever housing starts are in a given year the number of manufactured homes bought is between 15%-30% of that number. Tends to be closer to 30% when the economy is bad and closer to 15% when the economy is doing better, counter-cyclical for obvious reasons. In absolute numbers the MH industry has shipped in the range 250k-500k manufactured homes per year. You can figure 20% as a normalized number.

The past several years have seen an unprecedented depression in the MH industry. There was bad lending in the late 90's and early 00's that bankrupted a bunch of lenders that made loans to buy trailer homes. That removed financing and was the first leg down. But then you had the whole subprime fiasco; a huge proportion of people who normally would be the ones buying trailer houses were now able to buy a regular house with no money down. That was the thing that really killed MH demand and kept it there for several years. MH shipments have been running at about 8% of total housing starts(half the bottom end of the historical range) and 130k units in absolute terms.

Last year there was a lengthy period of time where CAV was trading in the range of $3-$3.50/sh. This was a decent estimate of their liquidation value, plus they were running at about breakeven. So right there, without worrying about anything else(and not worrying about the stock price movement) you know you won't lose money. The stock might go down more but it's not real, a company that is not losing money is not worth less than its' liquidation value.

So you're not going to lose money, and you know the MH industry is severely depressed and shipments would have to go up 2x just to return to the bottom end of the historical range(and it seems pretty likely we will get back near the top of the range given what is currently going on). Furthermore, CAV mothballed a lot of factories when the downturn hit, they can at least double their output without incurring any meaningful capex.

If their output doubled CAV would earn somewhere around $15mm-$20mm. At $3 the market cap was about $55mm. So you are buying a company where your downside is basically nil, and the upside is that you are buying it at 2x-3x it's normalized earnings.

If you read analyst reports from the time you will see lots of detailed predictions about gross margins, industry shipments, market share, operating expenses, Katrina rebuilding, interest rates, predictions on the housing market, etc etc etc. A ton of data and lots of analysis went into making those predictions and coming up with their quarterly earnings estimates.

But you don't even need to know any of those things, you don't need to come up with any sort of estimates and you don't need to have much insight.

Calculating the liquidation value is just a math problem involving addition and subtraction. Then you can look at long range data going back four decades and see that the industry follows a very predictable trend and that we are currently in a period of time that is completely unprecedented. The only insight you need is that the dynamics of the MH industry have not been permanently altered, the unprecedented decline in the industry is a direct result of the unprecedented availability of no money down mortgages to just about anybody. You don't need to calculate when the industry will return to historical levels, you just need to know that it is going to. It's essentially impossible to say what MH shipments are going to be over the next few years but it's very likely they will be significantly higher.

When it's hard or impossible to predict exactly how something will happen then analysts tend to not focus on it, and instead focus on all the data points. Q2 shipments below estimates?? DOWNGRADE! When in reality that kind of data is irrelevant in the grand scheme. But coming up with complex models incorporating lots of variables seems smart and gives you precise numbers, while saying you have no idea what exactly is going to happen but are pretty sure the stock is very undervalued doesn't sound so smart and doesn't give you any precise numbers to hang your hat on. I guess you could say focusing on data makes you miss the forest for the trees and also makes you tend to feel too confident in your analysis because you have hard numbers to look at.(Incidentally, Berkshire is now the largest owner of MH capacity in the US, with all the assets being bought since 2003.)

Re: Too Much Information
 

hawk59,

Coach me plz?

Re: Too Much Information
 

this too much information thing is why I am more attracted to learning TA than FA.

FA scares me a bit because I don't even know how to proceed to find the right information. And some people will have more info than I do no matter what I do. In TA, you just have to open stockcharts.com, and you have the same information as everyone else. The interpretation is still very subjective, but I feel less disadvantaged.

I'm not the type of person who would call the CFO of a company to try to get some info. I'm more the tyoe to stay behind my screen, looking at some price data.

Re: Too Much Information
 

I definitely didn't expect this post to inspire TA vs. FA discussion, but I suppose it makes some sense.

kimchi, just because you use less data than other analysts doesn't imply your data is any better. So your "one piece of data" comment is sort of loaded. I'm a little confused regarding your thoughts that "[Price] can't be massaged or manipulated like most other forms of data." What forms of data do you feel can be manipulated and how are those manipulations different than possible price manipulations. Marc Andreessen paints a pretty clear picture of price manipulation in a recent blog post about the Sowood hedge fund collapse:

"Given what we were facing and our uncertain ability to meet margin calls [we were leveraged -- we used debt to double down on our bets to juice returns, common for this class of hedge funds], we sought other buyers for some or all of the positions [all our peers on Wall Street smelled "blood in the water" and drove down market prices even further]."

Lastly, why use "hourly,daily, weekly, monthly charts" for price? Why not chart something else, like spread or volume? There's a big difference between using fewer inputs and using the right inputs, and you seem to be suggesting that one implies the other, even more specifically, that the first implies the second.


blueman, can you post an example of me writing long paragraphs to disagree with you when you were right?


hawk, "coming up with complex models incorporating lots of variables seems smart and gives you precise numbers, while saying you have no idea what exactly is going to happen but are pretty sure the stock is very undervalued doesn't sound so smart." I gave a presentation for an investment club at NYU on November 4, 2005 pitching Wynn Resorts (WYNN) as an investment idea. This was only a few months after they had opened their first casino in Vegas and there was virtually no operating data available yet. I chose not to include a valuation (I did include some statistics like win/table/day and revenue/room/night). The audience pressed me for a specific value in Q&A. This was not unreasonable because nearly all presentations in this club ended with a DCF that produced a specific dollar value of the company. My rebuttal went something like this:

"Let's say I offer you a bar of gold. You know, for a fact, that it is 100% real gold. I hand you the bar of gold, which is about as big as your phone, but you are not allowed to weigh it or in any other way attempt to determine its mass, volume, etc. Then I tell you that I'll sell you the bar for $50. Now do you care what the hell it's worth?"

Wynn got voted in by like 3 votes (25-22 or so) [img]/images/graemlins/smile.gif[/img]

Re: Too Much Information
 

[ QUOTE ]


"Let's say I offer you a bar of gold. You know, for a fact, that it is 100% real gold. I hand you the bar of gold, which is about as big as your phone, but you are not allowed to weigh it or in any other way attempt to determine its mass, volume, etc. Then I tell you that I'll sell you the bar for $50. Now do you care what the hell it's worth?"


[/ QUOTE ]

Definition of what makes a great investment...

Re: Too Much Information
 

[ QUOTE ]
[ QUOTE ]


"Let's say I offer you a bar of gold. You know, for a fact, that it is 100% real gold. I hand you the bar of gold, which is about as big as your phone, but you are not allowed to weigh it or in any other way attempt to determine its mass, volume, etc. Then I tell you that I'll sell you the bar for $50. Now do you care what the hell it's worth?"


[/ QUOTE ]

Definition of what makes a great investment...

[/ QUOTE ]
I'm not following. Is this sarcastic, serious, something else?

Re: Too Much Information
 

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]


"Let's say I offer you a bar of gold. You know, for a fact, that it is 100% real gold. I hand you the bar of gold, which is about as big as your phone, but you are not allowed to weigh it or in any other way attempt to determine its mass, volume, etc. Then I tell you that I'll sell you the bar for $50. Now do you care what the hell it's worth?"


[/ QUOTE ]

Definition of what makes a great investment...

[/ QUOTE ]
I'm not following. Is this sarcastic, serious, something else?

[/ QUOTE ]
Probably sarcasm.

Also, "I hand you the bar of gold, which is about as big as your phone, but you are not allowed to weigh it or in any other way attempt to determine its mass, volume, etc."

If you hand me the gold, I can still guesstimate its weight, look at what the market is paying per ounce, and determine if $50 is a bargain or not. So of course I care how much it is worth. Maybe the analogy was just a figure of speech and wasn't meant to be taken seriously? [img]/images/graemlins/confused.gif[/img]

Re: Too Much Information
 

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]


"Let's say I offer you a bar of gold. You know, for a fact, that it is 100% real gold. I hand you the bar of gold, which is about as big as your phone, but you are not allowed to weigh it or in any other way attempt to determine its mass, volume, etc. Then I tell you that I'll sell you the bar for $50. Now do you care what the hell it's worth?"


[/ QUOTE ]

Definition of what makes a great investment...

[/ QUOTE ]
I'm not following. Is this sarcastic, serious, something else?

[/ QUOTE ]

I'm baffled shouldn't this bar of gold be well into the five figure range? I mean if every case was this easy, I'd just take free money that people are handing out and sit around all day.

Re: Too Much Information
 

hawk59, DesertCat, and TAs,

Where can I go to find information on historical P/Es of different sectors in the stock market? After reading hawk59's post, I realized how the data can help me gauge a company's value amid Mr. Market's psychopathic behaviors.

Re: Too Much Information
 

Wow, sometimes I guess you really have to spell it out.

HERE WAS THE POINT OF WHAT I SAID

You can tell it worth a lot, and more importantly, a lot more than you're going to have to pay for it. You don't know exactly what it's worth and you could easily be off by 20% or more if someone asked you to fill in the value of the purchase. However, those limitations notwithstanding, it is still an obviously good idea to buy the gold.

Re: Too Much Information
 

To add the obligatory Buffett quote..
"It is better to be approximately right than precisely wrong."

Also, Gladwell's Blink seems to be very relevant to this discussion. He talks about how thin slicing can be more effective than people with more information in the examples of determining divorce, diagnosing heart attacks, hiring employees, etc. There are two things that Gladwell talk about that I think are important with investing. One is thin slicing exemplified with Buffett where he ignores the vast majority of information available to him. The other is snap judgements and the subconscious where experience investors like Soros thrive. He used to say he followed the "theory of reflexivity", but then later said that the theory was bogus and could be ignored. His son Robert said,

"My father will sit down and give you theories to explain why he does this or that. But I remember seeing it as a kid and thinking, Jesus Christ, at least half of this is [censored]. I mean, you know the reason he changes his position on the market or whatever is because his back starts killing him. It has nothing to do with reason. He literally goes into a spasm, and it's this early warning sign."

Re: Too Much Information
 

[ QUOTE ]
Wow, sometimes I guess you really have to spell it out.

HERE WAS THE POINT OF WHAT I SAID

You can tell it worth a lot, and more importantly, a lot more than you're going to have to pay for it. You don't know exactly what it's worth and you could easily be off by 20% or more if someone asked you to fill in the value of the purchase. However, those limitations notwithstanding, it is still an obviously good idea to buy the gold.

[/ QUOTE ]

Sorry, you know... Text has limitations with stretched examples and I kinda glossed over when I read through that Wynn stuff. Rereading it makes it very easy to understand.

Me: -1

Re: Too Much Information
 

[ QUOTE ]
I have always been amazed by the simplicity of Warren Buffett's approach. He does not have seas of analysts (I don't believe he may have any) nor does he do excessively deep diligence dives into companies.

[/ QUOTE ]
lol

I agree with the author's conclusion but his premises are kind of weak and seem to based on stuff he made up.

Re: Too Much Information
 

[ QUOTE ]
[ QUOTE ]
I have always been amazed by the simplicity of Warren Buffett's approach. He does not have seas of analysts (I don't believe he may have any) nor does he do excessively deep diligence dives into companies.

[/ QUOTE ]
lol

I agree with the author's conclusion but his premises are kind of weak and seem to based on stuff he made up.

[/ QUOTE ]
Not that I'm any kind of expert here, but I do know that Berkshire employees a tiny staff. I think less than 15 people work there. Compare that to other institutions that manage the kind of money that he does and it's a pretty glaring difference.

Re: Too Much Information
 

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
I have always been amazed by the simplicity of Warren Buffett's approach. He does not have seas of analysts (I don't believe he may have any) nor does he do excessively deep diligence dives into companies.

[/ QUOTE ]
lol

I agree with the author's conclusion but his premises are kind of weak and seem to based on stuff he made up.

[/ QUOTE ]
Not that I'm any kind of expert here, but I do know that Berkshire employees a tiny staff. I think less than 15 people work there. Compare that to other institutions that manage the kind of money that he does and it's a pretty glaring difference.

[/ QUOTE ]

I don't know about performance related statistics, but I have also found it very interesting how some $1-5 billion hedge funds and private equity funds employ a total of like 10 people and others employ 50. Would be interesting to see how this correlates.

Re: Too Much Information
 

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
I have always been amazed by the simplicity of Warren Buffett's approach. He does not have seas of analysts (I don't believe he may have any) nor does he do excessively deep diligence dives into companies.

[/ QUOTE ]
lol

I agree with the author's conclusion but his premises are kind of weak and seem to based on stuff he made up.

[/ QUOTE ]
Not that I'm any kind of expert here, but I do know that Berkshire employees a tiny staff. I think less than 15 people work there. Compare that to other institutions that manage the kind of money that he does and it's a pretty glaring difference.

[/ QUOTE ]

I don't know about performance related statistics, but I have also found it very interesting how some $1-5 billion hedge funds and private equity funds employ a total of like 10 people and others employ 50. Would be interesting to see how this correlates.

[/ QUOTE ]
The one with less employees but has the same capitalizations as other hedge funds obviously have a really really smart guy calling most, if not all, the investing shots.

For example, ESL and Berkshire don't have that many employees. The reasons are Eddie Lampert and Buffy.

Post deleted by Mat Sklansky
 

Re: Too Much Information
 

[ QUOTE ]
[ QUOTE ]


blueman, can you post an example of me writing long paragraphs to disagree with you when you were right?



[/ QUOTE ]

Sure. Pretty much every post you made in this thread

http://forumserver.twoplustwo.com/showfl...rt=all&vc=1

[/ QUOTE ]

I made three replies to you in that thread. Two were about moderating decisions (topic of the discussion and a previous ban that you said you agreed with) and one was a link and quote from an external news source (i.e. I cited someone else who disagreed with you).


As for the rest of your post, I don't really get it. A lot of people do the same sort of thing all the time: tell a group of people what big morons they are and then continue to interact with the group. Doesn't make much sense to me, but if it fulfills your desire to make sure everyone thinks you're an idiot then I guess it's working out.

Re: Too Much Information
 

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]


"Let's say I offer you a bar of gold. You know, for a fact, that it is 100% real gold. I hand you the bar of gold, which is about as big as your phone, but you are not allowed to weigh it or in any other way attempt to determine its mass, volume, etc. Then I tell you that I'll sell you the bar for $50. Now do you care what the hell it's worth?"


[/ QUOTE ]

Definition of what makes a great investment...

[/ QUOTE ]
I'm not following. Is this sarcastic, serious, something else?

[/ QUOTE ]

highly doubt thats sarcasm. you made a great point and DC si agreeing.

good analogy;
gold=stock
easy to value gold and while you acan't determine exactly what it is worth, you know for a fact (given your assumptions) that it is worth well more than $50.

Barron

Re: Too Much Information
 

[ QUOTE ]
Wow, sometimes I guess you really have to spell it out.

HERE WAS THE POINT OF WHAT I SAID

You can tell it worth a lot, and more importantly, a lot more than you're going to have to pay for it. You don't know exactly what it's worth and you could easily be off by 20% or more if someone asked you to fill in the value of the purchase. However, those limitations notwithstanding, it is still an obviously good idea to buy the gold.

[/ QUOTE ]

I wasn't being sarcastic. The best investments are easy decisions that don't require lots of data to support them. They are "obvious".

Re: Too Much Information
 

[ QUOTE ]
http://www.vcconfidential.com/2007/0...ch-inform.html

Some of you may have seen this from my shared feed. I thought it was pretty interesting and somewhat counter to the way a lot of people feel about investing. It specifically made me think about someone like Barron, that chooses to use tons of data.

One interesting note is that the author talks about not only investments in small, relatively inefficient markets (venture capital), but also uses Warren Buffett as an example.

Perhaps Homer Simpson was right, "Facts are meaningless. You could use facts to prove anything that's even remotely true!"

[/ QUOTE ]

the horse betting study was the best analogy in there and i think taleb used it as well.

what i do, fyi (evan et.al.) is to look at a ton of info and boil it down to the things that really matter to what i'm studying. like a filter.

for instance, when building a trading strategy that generates a signal, you need fo first have a ton of indicators based on economic data that determine that signal. the more info you use in indicators doesn't necessarily create a better system. using 10 indicators for a signal may be just as good or only marginally worse than using 50 or 100...just depends on the thing being created obviously.

the site you've shown does a good job of explaining why an overload of info could be bad and even make you think you're doing a better job than you would be able to do otherwise with less info.

good cite.

Barron

Re: Too Much Information
 

I merely skimmed the blog article, but the discussion here has me confused about the point of the article. Is the author trying to say that more information is distracting? Or is he making a point about the way people convey their analysis to others (i.e. people who sound smart aren't necessarily doing any better)? This discussion in this thread seems to be talking about the latter.

I think some people are misunderstanding the article, but Barron's recent reply seems in line with my understanding. I don't think it's the amount of information you examine or absorb that matters - in fact, at this stage, more is better. Success ultimately seems determined by the investor's ability to filter that information and determine the pieces that are useful/relevant to his or her analysis. Smarter people don't necessarily see more information as noise or a distraction - they are simply better at filtering it and making the intelligent decisions needed to process that information.

FWIW, I don't do any sort of stock analysis (I'm a standard passive index fund investor), but I think there's a subtle distinction between the quantity of information being the culprit of a poor analysis and the individual's ability to process that information as the culprit. I think it has to be the latter.

In conclusion, more is better unless you can't handle it.

Re: Too Much Information
 

[ QUOTE ]
In conclusion, more is better unless you can't handle it.

[/ QUOTE ]
I don't think that is a perfect conclusion. It is hard to determine if we are actually able to "handle" new information well. Often times the additional information you are getting seems helpful, but is just noise that can impair your decision making abilities. Once additional information is confirmed as just noise or is only very marginally beneficial, it is worth it to ignore it even if you can "handle it". To take an example from Blink, Dr. Brendan Reilly, the chairman of the Department of Medicine at Cook County Hospital in Chicago, instructed that the diagnosis of heart attacks be boiled down to 4 simple risk factors. The doctors were smart people who could use much additional and seemingly relevant information like if the patient has diabetes, a history of heart attacks, etc. However, once doctors used the simplified decision tree, it improved their ability to identify true heart attack patients by 70%. The key is that many years of evaluating and analyzing the evidence took place before the extra information was deemed as unhelpful.

Re: Too Much Information
 

If you knew for sure that information was "very marginally beneficial," why would you still choose to ignore it?

Re: Too Much Information
 

[ QUOTE ]
If you knew for sure that information was "very marginally beneficial," why would you still choose to ignore it?

[/ QUOTE ]
We're nitpicking here, but you might in cases where time is a factor, like when you are making a poker decision or, to use the blink example, diagnosing heart attacks. Also in cases where you are expending an exorbitant amount of resources for very little return, ie. mastering game theory to increase poker winnings or Buffett not having many analysts, etc.