"A Statistical Look at Jim Cramer's Skill Level"

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.

"A Statistical Look at Jim Cramer's Skill Level"

Postby Taylor Larimore » Mon May 27, 2013 3:15 pm

Bogleheads:

We have many posts warning about the danger of picking individual stocks--especially those recommended by media "experts." Boglehead Allen Roth illustrates the danger in this article about CNBC's stock guru, Jim Cramer.

A Statistical Look at Jim Cramer's Skill Level

Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle
User avatar
Taylor Larimore
Advisory Board
 
Posts: 20120
Joined: Tue Feb 27, 2007 9:09 pm
Location: Miami FL

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Mel Lindauer » Mon May 27, 2013 3:19 pm

Another great expose by Allan!

I'm sending in my money for Allan's REMARC hedge fund. It will be a sure winner (it will do the exact opposite of Cramer's buy and sell recommendations).

Allan will be a member of the Experts Panel at Bogleheads 12 in Philly in October. http://www.bogleheads.org/forum/viewtopic.php?p=1628287#p1628287
Best Regards - Mel | | Semper Fi
User avatar
Mel Lindauer
Moderator
 
Posts: 22335
Joined: Mon Feb 19, 2007 9:49 pm
Location: Daytona Beach Shores, Florida

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby OverTheHill » Mon May 27, 2013 3:51 pm

But, you forget how well he fills in dead air space every morning on CNBC. Anyone who claims to know as much as Cramer claims to know must be totally infatuated with themselves.
OverTheHill
 
Posts: 308
Joined: Wed Mar 27, 2013 5:58 pm
Location: North Carolina

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby mlebuf » Mon May 27, 2013 4:25 pm

Are people still taking Cramer seriously? I watched him once or twice about 10 years ago. He behaved like a middle aged man having a temper tantrum characteristic of a two year-old.

"No one ever went broke underestimating the intelligence of the American people."
- H.L. Mencken
Best wishes, | Michael | | Invest your time actively and your money passively.
User avatar
mlebuf
 
Posts: 1416
Joined: Tue Feb 20, 2007 9:27 pm
Location: Paradise Valley, Arizona

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby ThePrune » Mon May 27, 2013 4:35 pm

Jim's Mad Money show is just entertainment, like professional wrestling. Of course, there are people who think that wrestling isn't staged..... :shock:
Investment skill is often just luck in sheep's clothing.
User avatar
ThePrune
 
Posts: 802
Joined: Wed Nov 10, 2010 10:38 am
Location: Midland, MI

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby OverTheHill » Mon May 27, 2013 5:16 pm

ThePrune wrote:Jim's Mad Money show is just entertainment, like professional wrestling. Of course, there are people who think that wrestling isn't staged..... :shock:

Well, I can say that I watch about as much Mad Money as I watch professional wrestling, which is to say that I watch neither one, although it's possible that both shows share the same audience.
OverTheHill
 
Posts: 308
Joined: Wed Mar 27, 2013 5:58 pm
Location: North Carolina

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby tetractys » Mon May 27, 2013 5:33 pm

I think it was on NPR, on Market Place, where it was pointed out that a show of 'confidence' is much more important for market forecasters than 'accuracy.' That makes sense when one considers Cramer's minions marching to his money madness. -- Tet
User avatar
tetractys
 
Posts: 4198
Joined: Sat Mar 17, 2007 4:30 pm
Location: Along the Salish Sea

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby gatsby11 » Mon May 27, 2013 5:41 pm

While I agree with the premise of the article entirely, Jim Cramer is not one to listen to for stock evaluations, the author may want to look at his own statistical acumen as well:

Being a numbers guy, I couldn't resist calculating the odds of making four sell recommendations on what ends up being the four best performers out of 749 different stocks. Can we have a drum roll? The odds are 1 in 13.1 billion.



This calculation presumes that these were the only four calculations that Cramer made. This is clearly not the case. As a hypothetical counter example, let's say that Cramer made predictions about every stock, all stocks were equally likely to rise or fall, and Cramer had no inside information (meaning if markets are efficient, each stock has exactly 50/50 odds of underperforming/overperforming the index). If Cramer had made a prediction about every stock on the index, his odds of being wrong about the best performing four would be 2*2*2*2 = 1 in 16.

I think this is a lot closer to the reality of the situation, if clearly overstating the number of stock picks that Cramer made. While it's certainly great reason to not pay for the advice of an "expert", trying to use statistical exaggeration to drum up controversy and get page views is not my idea of excellence at one's job either, and unfortunately all too common in the media today.
Last edited by gatsby11 on Mon May 27, 2013 10:34 pm, edited 1 time in total.
gatsby11
 
Posts: 30
Joined: Sun May 26, 2013 2:58 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Dandy » Mon May 27, 2013 6:32 pm

I am not a Cramer fan. If you ignore all the hoopla you can see he is a smart, knowledgeable person who has some excellent reasons for his stock comments. But, that is the trap. It all sounds so smart and well thought out you think - yeah why that makes some sense -- and it does but that doesn't mean it translates into reliable stock market picks. Most of these pundits actually believe what they say.

It reminds me of the guys who tout college football picks. They all are knowledgeable and know that XYZ wins 80 of the home games, their opposition loses 70% of their away games etc, etc -- and they are lucky to get half the calls right. But they are so confident.

The secret is not to listen to these sirens who sound so good or your investment ship will end up on the rocks.
Dandy
 
Posts: 2388
Joined: Sun Apr 25, 2010 8:42 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby YttriumNitrate » Mon May 27, 2013 6:42 pm

I'm all for bashing Cramer, but the author of this article just cherry picked a couple of Cramer's predictions (from among hundreds if not thousands) and had the audacity to call his article a "statistical analysis"...
User avatar
YttriumNitrate
 
Posts: 724
Joined: Tue Mar 26, 2013 1:13 pm
Location: Midwest

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby rustymutt » Mon May 27, 2013 7:09 pm

Thanks for the link Taylor. I've got a neighbor whose fascinated with Cramer. I've tried to tell him that Cramer is merely a puppet of Wall Street, and used to
sensationalize the buying and selling of stocks. I've printed off the article Allen and will let him have a copy to read.
At the Very Least, Work Hard, Do Your Best, Know the Truth and the Facts and Always Be Honest!
User avatar
rustymutt
 
Posts: 2794
Joined: Sat Mar 07, 2009 1:03 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Allan Roth » Mon May 27, 2013 7:38 pm

YttriumNitrate wrote:I'm all for bashing Cramer, but the author of this article just cherry picked a couple of Cramer's predictions (from among hundreds if not thousands) and had the audacity to call his article a "statistical analysis"...


Not cherry picked at all. If you clicked on the links, you'd see I didn't use hindsight and actually invested prospectively with HP and Best Buy. I then looked at Cramer's calls on the only two stocks that did better than those two.

Cherry picking would have been stating "no, no, no, Bear Stearns is not in trouble, Apple is a temporary sell off in December due to cap gains tax changes, Wachovia, etc. Finally, I do think that as part of an education mission, Cramer should publicly state how his picks actually do. I'll bet you even money (okay 100 to 1 money) that you can't call the four worst performing stocks over the next six months?
Allan Roth
 
Posts: 251
Joined: Thu Sep 11, 2008 1:47 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby mlewis » Mon May 27, 2013 7:41 pm

YttriumNitrate wrote:I'm all for bashing Cramer, but the author of this article just cherry picked a couple of Cramer's predictions (from among hundreds if not thousands) and had the audacity to call his article a "statistical analysis"...



Here's some more good analysis by said author.

http://www.cbsnews.com/8301-505123_162- ... s-s-p-500/
User avatar
mlewis
 
Posts: 182
Joined: Mon Nov 05, 2012 1:39 am

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Levett » Mon May 27, 2013 7:55 pm

And here I thought Bogleheads basically ignored noise.

So if that's the case why bother with Cramer, the entertainer?

You are preaching to the choir.

Lev
Levett
 
Posts: 3348
Joined: Fri Feb 23, 2007 3:10 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Allan Roth » Mon May 27, 2013 8:06 pm

gatsby11 wrote: each stock has exactly 50/50 odds of underperforming/overperforming the index). If Cramer had made a prediction about every stock on the index, his odds of being wrong about the best performing four would be 2*2*2*2 = 1 in 16.



2 stocks were sampled prospectively which had a 1 in 35,062 chance of being 2 of the worst 4 of 749 stocks. IF Cramer had made a forecast of all 749 stocks then the probability of 1 in 140,248. Since I noted Cramer had made other predictions the statistical likelihood is somewhere between 1 in 140,248 and 1 in 13 billion.
Allan Roth
 
Posts: 251
Joined: Thu Sep 11, 2008 1:47 pm

The Three Fund Portfolio

Postby Taylor Larimore » Mon May 27, 2013 8:07 pm

Allan wrote:
My advice is to deploy formula X by buying a low cost-broad total index fund. Do the same for international stocks as well.

Your "advice" is the equity part of the The Three Fund Portfolio.

Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle
User avatar
Taylor Larimore
Advisory Board
 
Posts: 20120
Joined: Tue Feb 27, 2007 9:09 pm
Location: Miami FL

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby gatsby11 » Mon May 27, 2013 8:51 pm

2 stocks were sampled prospectively which had a 1 in 35,062 chance of being 2 of the worst 4 of 749 stocks.


No, YOU chose two stocks with the benefit of hindsight. Cramer forecasted many, many more than 2 of the 749 stocks.

IF Cramer had made a forecast of all 749 stocks then the probability of 1 in 140,248. Since I noted Cramer had made other predictions the statistical likelihood is somewhere between 1 in 140,248 and 1 in 13 billion.


This is just downright wrong Allan. If he had made a call of either buy or sell on every stock, then every stock has 1/2 odds of being sell. Since he predicted every stock in this hypothetical scenario he would be guaranteed to have selected the top four performers, the only question would be whether he forcasted buy or sell. Therefore:

Stock 1 sell odds: 1/2
Stock 2 sell odds: 1/2
Stock 3 sell odds: 1/2
Stock 4 sell odds: 1/2

Odds of all four being sell: 1/16

Simple intuition should tell you there's clearly no way the odds are anywhere near as low as you're saying.
Last edited by gatsby11 on Mon May 27, 2013 9:29 pm, edited 1 time in total.
gatsby11
 
Posts: 30
Joined: Sun May 26, 2013 2:58 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Boglenaut » Mon May 27, 2013 9:13 pm

Allen Roth is one of my favorite columnists, and I certainly think Jim Cramer says a lot of totally wrong things with great confidence. But I think Allen should stay away from statistical calculations for now. :oops: :oops:
User avatar
Boglenaut
 
Posts: 2891
Joined: Mon Mar 23, 2009 8:41 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Allan Roth » Mon May 27, 2013 9:28 pm

gatsby11 wrote:
This is just downright wrong Allan. If he had made a call of either buy or sell on every stock, then every stock has 1/2 odds of being sell. Since he predicted every stock in this hypothetical scenario he would be guaranteed to have selected the top four performers, the only question would be whether he forcasted buy or sell. Therefore:

Stock 1 sell odds: 1/2
Stock 2 sell odds: 1/2
Stock 3 sell odds: 1/2
Stock 4 sell odds: 1/2

Odds of all four being sell: 1/16

Simple intuition should tell you there's clearly no way the odds are anywhere near as low as you're saying.


Overconfidence bias on your part. You would be correct IF the methodology were picking from prior forecasts. It was not though I think it will be difficult for you to re-frame now that you are convinced you are right. The odds of picking two stocks on 11/20/12 that I invested in and wrote about at that time that were two of the best 4 stocks of 749 was 1 in 35,052 . You are thinking that picking four stocks retroactively where each stock has a 50% of being a sell (1/2)^4 is 1 of 16 and you would be right if that were the methodology.

I know a great statistician by the name of John Allen Paulos if you are willing the make the bet worthwhile?
Allan Roth
 
Posts: 251
Joined: Thu Sep 11, 2008 1:47 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Allan Roth » Mon May 27, 2013 9:30 pm

Boglenaut wrote:Allen Roth is one of my favorite columnists, and I certainly think Jim Cramer says a lot of totally wrong things with great confidence. But I think Allen should stay away from statistical calculations for now. :oops: :oops:


Thanks Boglenaut. I'm willing to bet it's not 1/16 if I have any takers?
Allan Roth
 
Posts: 251
Joined: Thu Sep 11, 2008 1:47 pm

Too Bad Jim Cramer didn't really commit professional suicide

Postby arcticpineapplecorp. » Mon May 27, 2013 9:31 pm

Why wasn't Cramer investigated for his 'treatise on manipulating stock prices' (see link below)?

http://dealbook.nytimes.com/2007/03/23/ ... -the-line/

"In a long examination of the clip, Mr. Blodget — who has been barred from the securities industry for life as part of a settlement over his stock research — predicts that Mr. Cramer’s career might be over. “Can CNBC really say nothing when one of its most visible employees urges investors to use the network to engage in behavior that is questionable to say the least?” he asked.

Mr. Blodget concludes:

It is certainly possible that the whole thing will pass: Cramer has been very effective at stifling or weathering critics. It is also possible, however, that Jim Cramer has committed professional suicide."

Too bad Mr. Blodget wasn't right.
The goal of investing is not to beat the market, but rather to beat inflation.
User avatar
arcticpineapplecorp.
 
Posts: 383
Joined: Tue Mar 06, 2012 10:22 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby abuss368 » Mon May 27, 2013 9:52 pm

Taylor Larimore wrote:Bogleheads:

We have many posts warning about the danger of picking individual stocks--especially those recommended by media "experts." Boglehead Allen Roth illustrates the danger in this article about CNBC's stock guru, Jim Cramer.

A Statistical Look at Jim Cramer's Skill Level

Best wishes.
Taylor



Hi Taylor,

Thank you for posting the link to that article.

One word: incredible!

I really do not miss the individual stock picking days.

Keep investing simple. The Three Fund Portfolio does just that.
John C. Bogle: "You simply do not need to put your money into 8 different mutual funds!" | | Disclosure: Three Fund Portfolio + REITs
User avatar
abuss368
 
Posts: 6584
Joined: Mon Aug 03, 2009 3:33 pm
Location: At the Beach

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby LadyGeek » Mon May 27, 2013 9:53 pm

In reference to some earlier posts - As a reminder:

We expect this forum to be a place where people can feel comfortable asking questions and where debates and discussions are conducted in civil tones... Attacks on individuals, insults, name calling, trolling, baiting or other attempts to sow dissension are not acceptable.

The forum policy extends to Mr. Cramer.
To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.
User avatar
LadyGeek
Site Admin
 
Posts: 19656
Joined: Sat Dec 20, 2008 6:34 pm
Location: Philadelphia

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby mlewis » Mon May 27, 2013 10:24 pm

Allan Roth wrote:
gatsby11 wrote:
This is just downright wrong Allan. If he had made a call of either buy or sell on every stock, then every stock has 1/2 odds of being sell. Since he predicted every stock in this hypothetical scenario he would be guaranteed to have selected the top four performers, the only question would be whether he forcasted buy or sell. Therefore:

Stock 1 sell odds: 1/2
Stock 2 sell odds: 1/2
Stock 3 sell odds: 1/2
Stock 4 sell odds: 1/2

Odds of all four being sell: 1/16

Simple intuition should tell you there's clearly no way the odds are anywhere near as low as you're saying.


Overconfidence bias on your part. You would be correct IF the methodology were picking from prior forecasts. It was not though I think it will be difficult for you to re-frame now that you are convinced you are right. The odds of picking two stocks on 11/20/12 that I invested in and wrote about at that time that were two of the best 4 stocks of 749 was 1 in 35,052 . You are thinking that picking four stocks retroactively where each stock has a 50% of being a sell (1/2)^4 is 1 of 16 and you would be right if that were the methodology.

I know a great statistician by the name of John Allen Paulos if you are willing the make the bet worthwhile?


With respect to Cramer, I think Gatsby is closer to being correct. Cramer of course made many, many predictions, so the odds of him telling his viewers to sell 2 stocks that went on to soar are not so long.
However, Allan Roth's long odds have more to do with himself. It is more statistically interesting that Roth decided to act opposite to 2 specific stock picks that happened to be so wrong.
User avatar
mlewis
 
Posts: 182
Joined: Mon Nov 05, 2012 1:39 am

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby gatsby11 » Mon May 27, 2013 10:32 pm

Overconfidence bias on your part.


Uh, no. It's just math. In the scenario I laid out the odds are 1/16, objectively.

The odds of picking two stocks on 11/20/12 that I invested in and wrote about at that time that were two of the best 4 stocks of 749 was 1 in 35,052 .


Yes, if on 11/20/12 you looked at those two picks alone and said they were be sells and they turned out as four of the best out of 749 on a set date that you predicted beforehand, you would have very long odds. But the odds are not what you say. Here are the real odds:

4 out of 749 + 3 out of 748

OR

187.25*249.33= 1 in 46,687

The problem is this is not what you did. You are looking back retroactively, noticing that Netflix and Best Buy were high at a time you had not predicted in advance, and used this cherry-picked data rather than all of the other forecasts Cramer has made.

Say instead that Cramer had made 100 picks during a month that you've picked in advance and you've also picked an end date for evaluation in advance. That means he made 50 sell picks. So ~1/16 of all of the stocks he's picked as sell. The odds that he would have picked 2 of the 4 worst performing stocks over that time period would be roughly:

1/16 * 1/16 = 1 in 256

Of course your mistake is much worse than this. You haven't selected an end date for evaluation in advance. 3 months ago Netflix and Best Buy wouldn't have been 2 of the top 4. In another 6 months they likely won't both be at the top either. So you picked an end date that suited you, because that's when the picks looked worst. That's not the same as picking the end date in advance, as you are fitting the data to your story and therefore have multiple data sets to pick from, drastically reducing how long the odds are. Even with just two data sets instead of one your odds are reduced by half.

And although you hint at it in your article's introduction I don't see anywhere you actually predicted this in advance. Assuming you didn't your mistake is compounded even further. Cramer's career covers a huge sample size of predictions. Let's no go that far though, we'll just look at one year. If we kept our earlier sample of 100 picks per month for 12 months, all of a sudden it's 12 times more likely that Cramer would have made such poor choices in the past year. That give us roughly a 1 in 20 chance of doing so. Clearly, Cramer has made many predictions and you could have written about any of them. The fact that you chose these two is no fluke, you saw that they were performing so well and therefore chose to write about them. If I make 2,000 predictions in my career, the odds that there will eventually be two poor predictions clustered closely together is fairly high.

This is all completely disregarding the fact that being in the top 4 rather than top 2 or top 10 or top 50 is another completely arbitrary number which you've chosen only in hindsight because it makes your data look the best.

I know a great statistician by the name of John Allen Paulos if you are willing the make the bet worthwhile?


By all means, define your terms. If we're betting how likely it is that Cramer would make such poor picks without selecting two stocks in advance and setting an end evaluation date in advance while also defining the data bin for being a "top performer" in advance though, the odds that you've stated in your article are off, and it's by orders of magnitude.
gatsby11
 
Posts: 30
Joined: Sun May 26, 2013 2:58 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Allan Roth » Mon May 27, 2013 10:47 pm

gatsby11 wrote:
Overconfidence bias on your part.

Uh, no. It's just math. In the scenario I laid out the odds are 1/16, objectively.



Not true in the scenario I wrote about - I invested int HPQ and BBY on 11/20/12 and did not know how it would turn out and this is stated in the article on how he made me money. I had to bet on his poor stock picking ahead of time and did not randomly pick stocks on 5/23/13.

To put this simply, you give me two stocks today in the Wilshire Large cap index which you can pull from Cramer or any other sources you want. If, six months later, they are two of the 4 best performers, I'll give you $16, but if you are wrong, you give me $1, which is 4x the 1 in 4 chance (1/2 x 1/2) you are claiming.

I would not recommend you take this wager as the math is not on your side.
Allan Roth
 
Posts: 251
Joined: Thu Sep 11, 2008 1:47 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Boglenaut » Mon May 27, 2013 10:50 pm

Allan Roth wrote:
Boglenaut wrote:Allen Roth is one of my favorite columnists, and I certainly think Jim Cramer says a lot of totally wrong things with great confidence. But I think Allen should stay away from statistical calculations for now. :oops: :oops:


Thanks Boglenaut. I'm willing to bet it's not 1/16 if I have any takers?



To be honest, your methodology was so flawed it really is hard to even start. I don't think anyone expects MoneyWatch articles to be at the same level as a peer reviewed academic paper, but you should use the words "A Statistical look" a bit more cautiously. You cannot just cherry-pick a few calls after the fact and draw the conclusions you did.

I just did a Google search and found this.

http://caps.fool.com/player/trackjimcramer.aspx

Accuracy: 45.13% (33rd percentile)

It seems a bit more objective in that it shows all buy/sell and has a time frame, and has a much larger sample. It seems to track 2373 Jim Cramer calls since 05/15/06.

It's bedtime, so I didn't have time to study it in detail.
User avatar
Boglenaut
 
Posts: 2891
Joined: Mon Mar 23, 2009 8:41 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby gatsby11 » Mon May 27, 2013 11:04 pm

To put this simply, you give me two stocks today in the Wilshire Large cap index which you can pull from Cramer or any other sources you want. If, six months later, they are two of the 4 best performers, I'll give you $16, but if you are wrong, you give me $1, which is 4x the 1 in 4 chance (1/2 x 1/2) you are claiming.

I would not recommend you take this wager as the math is not on your side.


That much is clear. That's also quite obviously not what I said.
gatsby11
 
Posts: 30
Joined: Sun May 26, 2013 2:58 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Random Musings » Mon May 27, 2013 11:05 pm

Jim Cramer has a reasonably high skill level when it comes to making money for himself.

In fact, much higher than most people.

RM
User avatar
Random Musings
 
Posts: 5035
Joined: Thu Feb 22, 2007 5:24 pm
Location: Pennsylvania

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Boglenaut » Mon May 27, 2013 11:13 pm

To be fair to Cramer, in the above link I cited:

    Average Pick Score is a player's total score divided by the total number of picks (active and closed). It represents the player's average return after subtracting out the market's performance.

    Average Pick Score: +1.86

So that's not bad. But if it were a fund, fees and expenses would eat much of it, with the additional risk associated with less diversity (even if he could maintain it). I wouldn't buy it.
User avatar
Boglenaut
 
Posts: 2891
Joined: Mon Mar 23, 2009 8:41 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Brian2d » Mon May 27, 2013 11:14 pm

How do Jim Cramer's picks compare with Cosmo Kramer's?
Brian2d
 
Posts: 157
Joined: Thu Jan 20, 2011 10:04 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Allan Roth » Mon May 27, 2013 11:17 pm

gatsby11 wrote:
To put this simply, you give me two stocks today in the Wilshire Large cap index which you can pull from Cramer or any other sources you want. If, six months later, they are two of the 4 best performers, I'll give you $16, but if you are wrong, you give me $1, which is 4x the 1 in 4 chance (1/2 x 1/2) you are claiming.

I would not recommend you take this wager as the math is not on your side.


That much is clear. That's also quite obviously not what I said.


What you said is that my piece had a 1/16 chance even though I invested in HPQ and BBY as soon as I read Cramer's sells on 11/20/12. My wager proposal was to show that your 1/16 calculation was far off the mark for the scenario I wrote about. It would be true for the probability of getting four heads in a row on a series of coin flips. "Simple intuition" can often be wrong.
Last edited by Allan Roth on Mon May 27, 2013 11:28 pm, edited 2 times in total.
Allan Roth
 
Posts: 251
Joined: Thu Sep 11, 2008 1:47 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Allan Roth » Mon May 27, 2013 11:19 pm

Random Musings wrote:Jim Cramer has a reasonably high skill level when it comes to making money for himself.

In fact, much higher than most people.

RM


That much is true but let's not forget what Cramer does to keep the stock market efficient for us indexers.
Allan Roth
 
Posts: 251
Joined: Thu Sep 11, 2008 1:47 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Random Musings » Mon May 27, 2013 11:28 pm

Allan Roth wrote:
Random Musings wrote:Jim Cramer has a reasonably high skill level when it comes to making money for himself.

In fact, much higher than most people.

RM


That much is true but let's not forget what Cramer does to keep the stock market efficient for us indexers.


It is comforting to know that Cramer is taking one for the team. :oops:

RM
User avatar
Random Musings
 
Posts: 5035
Joined: Thu Feb 22, 2007 5:24 pm
Location: Pennsylvania

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby beardsworth » Mon May 27, 2013 11:41 pm

Mel Lindauer wrote:Another great expose by Allan!

I'm sending in my money for Allan's REMARC hedge fund. It will be a sure winner (it will do the exact opposite of Cramer's buy and sell recommendations).


I don't think a fund called REMARC would be right for me. :)
beardsworth
 
Posts: 1557
Joined: Fri Jun 15, 2007 5:02 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Levett » Tue May 28, 2013 6:53 am

My compliments to Lady Geek who practices what this site ostensibly preaches:

"Attacks on individuals, insults, name calling, trolling, baiting or other attempts to sow dissension are not acceptable."

Would that the policy were implemented with more regularity and without exception.

If Bogleheads have a positive message, let it remain positive.

Lev
Levett
 
Posts: 3348
Joined: Fri Feb 23, 2007 3:10 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby YttriumNitrate » Tue May 28, 2013 7:13 am

Allan Roth wrote:]I'll bet you even money (okay 100 to 1 money) that you can't call the four worst performing stocks over the next six months?


At 100 to 1 I'd take that bet, even at 20 to 1 the odds would be in my favor. Of course, like Cramer I get to make a lot of calls, a LOT of calls. Unlike Cramer who tries to predict the future with insight or a crystal ball, my calls are based solely on the ticker symbol chosen by the company. I hereby declare those stocks starting with the first 13 letters of the alphabet to be strong buys while those starting with the last 13 letters of the alphabet to be strong sells.

Now, how much are we wagering?
User avatar
YttriumNitrate
 
Posts: 724
Joined: Tue Mar 26, 2013 1:13 pm
Location: Midwest

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby rustymutt » Tue May 28, 2013 9:01 am

The problem with Cramer is that there isn't a benchmark that is used to track his picks. If fact he's got nothing to compare himself to. He dwells on confusion and hype. Sure he guesses right at times, but he's wrong just as many times. He's a Wall Street poster child, and no intelligent investor should be using his media hype as advise.
At the Very Least, Work Hard, Do Your Best, Know the Truth and the Facts and Always Be Honest!
User avatar
rustymutt
 
Posts: 2794
Joined: Sat Mar 07, 2009 1:03 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Allan Roth » Tue May 28, 2013 10:35 am

YttriumNitrate wrote:
Allan Roth wrote:]I'll bet you even money (okay 100 to 1 money) that you can't call the four worst performing stocks over the next six months?


At 100 to 1 I'd take that bet, even at 20 to 1 the odds would be in my favor. Of course, like Cramer I get to make a lot of calls, a LOT of calls. Unlike Cramer who tries to predict the future with insight or a crystal ball, my calls are based solely on the ticker symbol chosen by the company. I hereby declare those stocks starting with the first 13 letters of the alphabet to be strong buys while those starting with the last 13 letters of the alphabet to be strong sells.

Now, how much are we wagering?


We are on with this wager. Using the same methodology in this piece, give me two stocks today that will be in the top 4 performers of the 749 in the Wilshire Large Cap index. Then, I'll give you the benefit of the doubt and allow you to assign an equal number of buys and sells of the remaining 747 stocks (you pick which one gets the extra). If six months later, your 2 stocks chosen today and both of your random sells are in the top four, I pay you 100 to 1. You pick the amount of the bet and I look forward to hearing from you directly for this wager and one of us can post on this site.

FYI, I teach a behavioral finance class at The University of Denver where I use some simple math problems to illustrate heuristic biases - mental short-cuts that lead us to gross errors in solving. I understand why you think the odds of the events in my piece are 1/16 but I caution you to reconsider accepting this bet.
Allan Roth
 
Posts: 251
Joined: Thu Sep 11, 2008 1:47 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Boglenaut » Tue May 28, 2013 10:49 am

rustymutt wrote:The problem with Cramer is that there isn't a benchmark that is used to track his picks.


Didn't you see my link above? What more could you want?

http://caps.fool.com/player/trackjimcramer.aspx

I think short of a full-blown academic study, this is about as close as you are going to get.

Accuracy: 45.13%
Average Pick Score: +1.93
Average Pick Rating: 3 stars out of 5.

So about so-so.
User avatar
Boglenaut
 
Posts: 2891
Joined: Mon Mar 23, 2009 8:41 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Boglenaut » Tue May 28, 2013 10:59 am

Allan Roth wrote:
FYI, I teach a behavioral finance class at The University of Denver where I use some simple math problems to illustrate heuristic biases - mental short-cuts that lead us to gross errors in solving. I understand why you think the odds of the events in my piece are 1/16 but I caution you to reconsider accepting this bet.


Actually, the odds of what happened in your article happening is 100%. You basically said "Given Netflix skyrocketed and given I know Jim Cramer called this a sell, what are the odds that Jim Cramer called it a sell and Netflix skyrocketed?". 100%

I like the Motley Fool methodology much better. They are tracking 100% of the calls, have a large sample size, have defined conditions for success/failure, established metrics, defined time period. I could nitpick parts of it, but overall it seems more objective.
User avatar
Boglenaut
 
Posts: 2891
Joined: Mon Mar 23, 2009 8:41 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby investingdad » Tue May 28, 2013 11:03 am

Allan Roth wrote:
YttriumNitrate wrote:
Allan Roth wrote:]I'll bet you even money (okay 100 to 1 money) that you can't call the four worst performing stocks over the next six months?


At 100 to 1 I'd take that bet, even at 20 to 1 the odds would be in my favor. Of course, like Cramer I get to make a lot of calls, a LOT of calls. Unlike Cramer who tries to predict the future with insight or a crystal ball, my calls are based solely on the ticker symbol chosen by the company. I hereby declare those stocks starting with the first 13 letters of the alphabet to be strong buys while those starting with the last 13 letters of the alphabet to be strong sells.

Now, how much are we wagering?


We are on with this wager. Using the same methodology in this piece, give me two stocks today that will be in the top 4 performers of the 749 in the Wilshire Large Cap index. Then, I'll give you the benefit of the doubt and allow you to assign an equal number of buys and sells of the remaining 747 stocks (you pick which one gets the extra). If six months later, your 2 stocks chosen today and both of your random sells are in the top four, I pay you 100 to 1. You pick the amount of the bet and I look forward to hearing from you directly for this wager and one of us can post on this site.

FYI, I teach a behavioral finance class at The University of Denver where I use some simple math problems to illustrate heuristic biases - mental short-cuts that lead us to gross errors in solving. I understand why you think the odds of the events in my piece are 1/16 but I caution you to reconsider accepting this bet.


May I suggest that such a wager see the proceeds donated to the winner's preferred (non political) charity in the name of Bogleheads?
investingdad
 
Posts: 613
Joined: Fri Mar 15, 2013 11:41 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby YttriumNitrate » Tue May 28, 2013 12:01 pm

Allan Roth wrote:We are on with this wager. Using the same methodology in this piece, give me two stocks today that will be in the top 4 performers of the 749 in the Wilshire Large Cap index.


I think we’re misunderstanding the rules of the wager. I’m willing to play the role of Cramer in the article (i.e., making hundreds of blind predictions about the future) but not the role of Allan Roth in the article (picking just two of the stocks).
User avatar
YttriumNitrate
 
Posts: 724
Joined: Tue Mar 26, 2013 1:13 pm
Location: Midwest

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby hicabob » Tue May 28, 2013 12:21 pm

Everyone seems to love to bash Cramer. Having done some work in Cramer's studio back in the Cramer /Kudlow days I have met and watched JC up close. The people that work around him (with the exception of Kudlow) appeared to adore him! He had a reputation as a very decent and kind man and he is easy to work with (and he brings bagels in!) . As with all commercial TV, the purpose of his show is to maximize viewers so maximizing ad revenues. Of course, anyone that confuses him with a financial advisor is incredibly naive to say the least.
hicabob
 
Posts: 1748
Joined: Fri May 27, 2011 6:35 pm
Location: cruz

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Allan Roth » Tue May 28, 2013 12:29 pm

YttriumNitrate wrote:
Allan Roth wrote:We are on with this wager. Using the same methodology in this piece, give me two stocks today that will be in the top 4 performers of the 749 in the Wilshire Large Cap index.


I think we’re misunderstanding the rules of the wager. I’m willing to play the role of Cramer in the article (i.e., making hundreds of blind predictions about the future) but not the role of Allan Roth in the article (picking just two of the stocks).


I think you are misunderstanding the article. Cramer gave two predictions on 11/20/12 and I acted on both immediately after reading. These are the same rules as the article itself and I'm offering to let you play the same role as Cramer in the article and giving you a 100 to 1 payoff. You must give me two stocks today that will be in the top four in six months. By allowing you to predict all of the remaining 747 stocks, you are actually getting more generous rules than Cramer had in the article. I strongly recommend you not accept this wager but am willing to make it. If you do, I suggest the proceeds go to the Bogleheads. Please contact me directly if you want to proceed.
Allan Roth
 
Posts: 251
Joined: Thu Sep 11, 2008 1:47 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby EyeYield » Tue May 28, 2013 12:44 pm

What the heck is going on in the Boglehood?? I expect an occasional speculation thread, but gambling on speculation? Please.

Lets call the show Mad Money what it is: A tv show. A tv show that is sponsored by brokerage firms who make money from trading costs.

Cramer is doing a great job apparently, since the advertisers don't seem to be going away and his show is holding it's own in it's time slot.

His job is to get people to trade; and then trade some more. The viewers do, we don't. Is there really anything more that needs to be said?

If there's going to be casino action here, shouldn't there be a separate section for it? Why don't we just set up a roulette wheel and replace the numbers with ticker symbols and the colors with indexes then have at it? At least that would be actionable.

Sorry, digressing fast..... :oops:
"The stock market is a giant distraction from the business of investing." - Jack Bogle
EyeYield
 
Posts: 307
Joined: Tue Sep 18, 2012 7:43 pm
Location: Extremistan

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Buysider » Tue May 28, 2013 12:52 pm

Just because one might not think they can beat the market, does not mean one should not have an opinion about the price of a security. Whether for just for fun (like most folks with Cramer), or for real (AstraZeneca just bought Omthera OMTH today at a 100% premium to the prior market price), prices are just the meeting of opinions of buyers and sellers. Lots of people have opinions, you may not agree with how or why they have them, but they are setting the price you are paying to buy or receiving when selling a security...

Companies in the S&P 500 are making decisions about the price of other companies' securities, and while you might hope they are better informed than Cramer, you might be disappointed.
Buysider
 
Posts: 608
Joined: Sun May 18, 2008 10:36 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby mlewis » Tue May 28, 2013 1:04 pm

Allan Roth wrote:
YttriumNitrate wrote:
Allan Roth wrote:We are on with this wager. Using the same methodology in this piece, give me two stocks today that will be in the top 4 performers of the 749 in the Wilshire Large Cap index.


I think we’re misunderstanding the rules of the wager. I’m willing to play the role of Cramer in the article (i.e., making hundreds of blind predictions about the future) but not the role of Allan Roth in the article (picking just two of the stocks).


I think you are misunderstanding the article. Cramer gave two predictions on 11/20/12 and I acted on both immediately after reading. These are the same rules as the article itself and I'm offering to let you play the same role as Cramer in the article and giving you a 100 to 1 payoff. You must give me two stocks today that will be in the top four in six months. By allowing you to predict all of the remaining 747 stocks, you are actually getting more generous rules than Cramer had in the article. I strongly recommend you not accept this wager but am willing to make it. If you do, I suggest the proceeds go to the Bogleheads. Please contact me directly if you want to proceed.


The role of cramer was to make predictions on many stocks. The role of Mr. Roth was to somehow choose the two predictions that were way off. Although he did note the time period and write about them before-hand, I didn't notice that he set 6-month timeframe in advance, as others have noted. My guess is that since there are such long odds of Allen choosing the 2 picks that were so far off, he probably found some time to watch more of cramers picks over multiple timeframes and could have written many different articles.
User avatar
mlewis
 
Posts: 182
Joined: Mon Nov 05, 2012 1:39 am

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby Allan Roth » Tue May 28, 2013 1:22 pm

mlewis wrote:
The role of cramer was to make predictions on many stocks. The role of Mr. Roth was to somehow choose the two predictions that were way off. Although he did note the time period and write about them before-hand, I didn't notice that he set 6-month timeframe in advance, as others have noted. My guess is that since there are such long odds of Allen choosing the 2 picks that were so far off, he probably found some time to watch more of cramers picks over multiple timeframes and could have written many different articles.


Events happened as described in the piece and I assure you I don't often watch Cramer. I invested in the two stocks the day I read about his sell recommendations. I then looked at his calls on the only two stocks that bested those two and they turned out to be sell recommendations as well. Your guess that I was watching more of Cramer's picks isn't right but I do appreciate you characterizing it as a guess rather than fact. As far as the six month period, what period would you suggest?
Allan Roth
 
Posts: 251
Joined: Thu Sep 11, 2008 1:47 pm

Re: "A Statistical Look at Jim Cramer's Skill Level"

Postby johnubc » Tue May 28, 2013 1:31 pm

although the article seems to be factual, it is not scientific. I can look back at my own predictions and see where I should have bought and sold stocks. What would have been more interesting is if they were able to do an expose on the Action Alerts Plus newsletter where Jim actually (supposedly) has his portfolio.
johnubc
 
Posts: 347
Joined: Wed Jan 06, 2010 7:54 am

Next

Return to Investing - Theory, News & General

Who is online

Users browsing this forum: Chan_va, Google [Bot], jj2323us and 40 guests