Printer Friendly

"You! Me! Let's try to make some mo-ney!!!".

Executive Summary

Since March 14, 2005, nearly every business day at 6 PM Eastern Standard Time (EST) there is an hour-long show on the cable financial network CNBC. The program, called Mad Money, is billed as a briefing on personal investment. Jim Cramer, the energetic host, offers his "buy" and "sell" recommendations of companies in an effort to help the audience to "... try to make money." We analyze recommendations made on Jim Cramer's show along two lines of inquiry. First, we estimate the immediate impact of Cramer's recommendations by considering the market reaction subsequent to a broadcast. Second, we examine the effect of his investment advice on the stock price of Mad Money picks, and on the direct competitors of the recommended companies.

Our sample consists of episodes broadcast between August 2008 and December 2008. Mad Money shows are aired twice nightly on CNBC five days a week, first at 6 PM EST and then repeated at 11 PM EST. The data were taken from official recaps posted at http://www.thestreet.com, a financial website founded by Jim Cramer, and is restricted to companies listed on the NYSE and AMEX exchanges as well as NASDAQ market. We identify Cramer's picks as either "buy" or "sell." This simple binary categorization eliminates distinguishing between weak and strong recommendations (positive or negative) and removes the subjectivity that would be involved in such a transcription process.

Next, we separate picks by the party who initiated the ticker question: caller or non-caller. During each segment -- such as "Lightning round," "Sudden death," and "Are you diversified?" -- viewers call into the program and ask Jim Cramer for his opinion on particular stocks. He discusses these callers' picks very briefly. Therefore an average five-second recommendation during the "Lightning round" is unlikely to convey the same information as Cramer's five-minute recommendation during the opening segment. Furthermore, the bulk of the tickers mentioned on the program represent questions from callers, and Jim Cramer professes no previous knowledge that these tickers will be investigated. Therefore, his responses to callers' inquiries may not represent his stock-picking ability.

Cramer devotes the majority of each broadcast to discussing his "picks" -- stocks that he has complete discretion to include in the program. As a result, those "picks" are reflective of his stock-picking ability. We then identify competitors of in-depth Cramer "buy" and "sell" picks using the four-digit SIC code (the Standard Industrial Classification which classifies industries by a four digital code), and each firm's size, and obtain their daily pricing information over the sample period. Stock prices were collected for both Cramer's selections and for the

competitors, on the date they were recommended and five days after the recommendation date.

Results in our study indicate that the prices of Jim Cramer's buy picks experienced significantly lower prices after the buy recommendations. The competitors of the buy picks experienced even worse stock performance after the recommendations. The price differences between the two groups are, on average, statistically significant. According to our results, potential investors will not profit from Jim Cramer's buy picks. However, results also show that Cramer's picks outperformed the competitors in the same industry.

Regarding sell recommendations, evidence in this study shows that prices of Cramer's sell picks do experience lower stock prices, and the declines in prices are significantly lower than the price as of the recommendation date. The same conclusion can be drawn for the competitors. The differences between sample and competitors, however, are inconclusive.

1. Introduction

Since March 14, 2005, nearly every business day at 6 pm EST Jim Cramer has been hosting Mad Money, an hour-long show providing investment advice. Presented on the cable financial network CNBC, the energetic host gives viewers his "buy" and "sell" recommendations. Cramer professes no insider knowledge and encourages his audience -600,000 viewers daily -- to research each firm's earnings estimates before investing. It is safe to presume that Cramer's picks -- at least over a short horizon -- should be reasonably correct. To determine their accuracy, a "buy" recommendation on the show, followed by a swell of buy orders and spikes in prices the next trading day, would provide instant proof. However, in August 2007, Barron's, a major financial newspaper, revealed that Cramer's stock picks consistently underperform.

Our research studies an interaction between stock prices and analyst recommendation literatures. The mass-media component of our paper examines the market reaction to stocks mentioned in financial news outlets (e.g., Barber and Loeffler, 1993; Greene and Smart, 1999; Liang, 1999; Busses and Green, 2002; Barber and Odean, 2006; Engelberg et al, 2007; Lim and Rosario, 2008). For example, Barber and Odean (2006) studied account activity at large retail brokerage firms and determined that individual investors trade around attention-grabbing events, and lose money by doing so. Engelberg et al. (2007; ESW) uncovered significant next-day returns for first-time positive recommendations by Cramer, returns which increased inversely with market capitalization. Lim and Rosario (2008) documented smaller excess returns subsequent to recommendations than ESW found. They attribute them to either the difference in sample selection or to the difference in measurement of excess returns.

The purpose of this paper is to examine whether potential investors can benefit from Jim Cramer's picks, both buy and sell recommendations. More specifically, we evaluate the market response immediately following a broadcast of Mad Money. We analyze recommendations made by Jim Cramer along two lines of inquiry. First, we estimate the immediate impact of Cramer's recommendations by the market reaction subsequent to a broadcast. Second, we examine the effect of his investment advice on the stock price of Mad Money picks and on the direct competitors of the recommended companies. Cohen and Frazzini (2006) argue that in the market in which all participants are potential recipients of news, investors underreact to news from economically linked firms. Neumann and Kenny's (2007) analysis of returns and trading volume around stock recommendations aired on the Mad Money program reveals investors' response to both "buy" and "sell" recommendations.

2. Data

Our sample consists of episodes broadcast between August 2008 and December 2008. (1) Mad Money shows are aired nightly on CNBC five days a week, first at 6 PM EST and then repeated at 11 PM EST. Characterized as a briefing on personal investment, the program is delivered with flair and infused with gimmicks that are intended to enhance the entertainment value of the show, with Jim Cramer offering his recommendations and criticisms of companies in an effort to, in his words, "make you money." The majority of each broadcast's running time consists of a discussion of firms he elects to talk about -- Cramer's picks. These are companies he has researched beforehand and has chosen to mention on the program. Cramer has complete discretion over the firms in this category. There are also smaller segments of the program called "Lighting Round," "Sudden Death," and "Are you Diversified?" During those segments, viewers call in to the program and ask Cramer for his opinion on particular stocks they are curious about. Caller picks are typically discussed only cursorily.

The data were taken from official recaps posted at http://www.thestreet.com, a financial website founded by Jim Cramer, and is restricted to companies listed on the NYSE and AMEX exchanges as well as the NASDAQ market. We checked results against independent recap websites. (2)

We identify picks as either "buy" or "sell." This simple binary categorization eliminates distinguishing between weak and strong recommendations (positive or negative) and removes the subjectivity that would be involved in such a transcription process. Next, we separate picks by the party who initiated the ticker question: caller or non-caller (Cramer).

During the short segments when viewers call in to ask Cramer's opinion on particular stocks, he discusses these 'callers' picks' very briefly. Therefore an average five-second recommendation during the "Lightning round" is unlikely to convey the same information as one of his own five-minute recommendations, which he discusses during the opening segment. The bulk of the tickers mentioned on the program represent questions from callers, and Jim Cramer professes no previous knowledge that those tickers will be discussed on a show. Therefore, his responses to callers' inquiries may not represent his stock-picking ability. It is worth mentioning that the majority of these tickers are often follow-ups to his previous "buy" or "sell" picks.

Cramer devotes the majority of each broadcast's time to discussing his "picks" -- stocks that he includes in the program completely at his own discretion, and which he has presumably investigated. As a result, these "picks" are reflective of his stock-picking ability.

Next, we identify competitors of in-depth Cramer "buy" and "sell" picks using the four-digit SIC code (the Standard Industrial Classification to classify industries by a four digital code) and the firm's size, and obtain their daily pricing information over the sample period. Stock prices at the recommendation date as well as five days after recommendation date were collected for both Cramer's selections sample and competitor firms.

Over 75% of Cramer's in-depth picks represent large cap stocks, responses to phone inquiries or both. If he is correct:

* a price change for an in-depth "buy" recommendation is likely to be positive, whereas

* a price change for an in-depth "sell" recommendation is likely to be negative.

Therefore, we anticipate spikes and valleys in the short-term price change of his picks due to a majority of his audience following his advice. The response, on average, of the picks' competitors is less easy to anticipate.

3. Results

The results in Exhibit 1, Panel A, show that, on average, 52.47% of the sample companies did experience a higher price one day after the "buy" recommendation.

According to Exhibit 1, Panel B, 64.60% of the companies experienced a lower price one day after the "sell" recommendation. On the other hand, about two-thirds of the recommended "sell" companies also experienced lower prices two, three, four, and five days after the recommendation date.

It is also noteworthy that over the sample period "buy" recommendations outnumber "sell" recommendations (223 vs. 101).
Exhibit 1
Descriptive statistics of price differences for sample
companies: "buy" and "sell" recommendations

Panel A. "Buy" recommendation   % of firms that         Std.
N=223                           experienced positive    Dev.
                                change in stock
                                price

  [P.sub.1] - [P.sub.0] *               52.47           0.50051

    [P.sub.2] - [P.sub.0]               39.91           0.49082

    [P.sub.3] - [P.sub.0]               41.70           0.49418

    [P.sub.4] - [P.sub.0]               42.60           0.49561

    [P.sub.5] - [P.sub.0]               39.82           0.49064

Panel B. "Sell" recommendation  % of firms that         Std.
N=101                           experienced positive    Dev.
                                change in stock
                                price

  [P.sub.1] - [P.sub.0] *               35.40           0.48133

    [P.sub.2] - [P.sub.0]               32.45           0.48439

    [P.sub.3] - [P.sub.0]               36.28           0.46352

    [P.sub.4] - [P.sub.0]               33.92           0.45923

    [P.sub.5] - [P.sub.0]               31.27           0.46756

Note: (*) P1-P0 = Price one day after the recommendation date minus
price on the recommendation date. If the price difference is
positive, it is coded as "1"; otherwise, it is coded as "0".


Next, we separate companies into two groups. Group 1 represents companies that experience price decreases after the recommendation day. Companies that experience price increases after recommendation day are placed in Group 2.

According to the non-parametric test results presented in Exhibit 2, Panel A, 60% of companies with "buy" recommendations experience price decreases (3) two, three, four, and five days after the recommendation has been "aired" on TV.

52% of companies given "buy" recommendations by Jim Cramer do exhibit higher prices, but only on the first day after his recommendations (4). In other words, contrary to Jim Cramer's predictions, results indicate that on average, stocks that Cramer gave "buy" recommendations experienced lower prices for the next few days after day 1.

On the other hand, a price drop (5) after "sell" recommendations takes place the day after the recommendation date. We observe that individual investors are more sensitive to "sell" recommendations vs. "buy" recommendations. On average, dose to 70% of "sell" recommendation companies experience price decreases the first five days after recommendation day. It appears that stocks with "sell" recommendations experience negative prices longer than one day, while a "buy" recommendation has less positive effect on stock prices after recommendation day.
Exhibit 2 Non-parametric binomial test stock price
change for sample of companies: Test proportion

Panel A. "Buy" recommendation N= 223
Prop                                     N

[P.sub.1]-[P.sub.0]                   Group 1

                                      Group 2

[P.sub.2]-[P.sub.0]                   Group 1

                                      Group 2

[P.sub.3]-[P.sub.0]                   Group 1

                                      Group 2

[P.sub.4]-[P.sub.0]                   Group 1

                                      Group 2

[P.sub.5]-[P.sub.0]                   Group 1

                                      Group 2

                                      Total

Panel B. "Sell lrecommendation N=101  N
(2-tailed)

[P.sub.1]-[P.sub.0]                   Group 1

                                      Group 2

[P.sub.2]-[P.sub.0]                   Group 1

                                      Group 2

[P.sub.3]-[P.sub.0]                   Group 1

                                      Group 2

[P.sub.4]-[P.sub.0]                   Group 1

                                      Group 2

[P.sub.5]-[P.sub.0]                   Group 1

                                      Group 2

                                      Total

Panel A. "Buy" recommendation N= 223  Observed       Test
Prop                                  Asymp.         Prop.

[P.sub.1]-[P.sub.0]                   106            0.48

                                      117            0.52

[P.sub.2]-[P.sub.0]                   134            0.60

                                      89             0.40

[P.sub.3]-[P.sub.0]                   130            0.58

                                      93             0.42

[P.sub.4]-[P.sub.0]                   128            0.57

                                      95             0.43

[P.sub.5]-[P.sub.0]                   133            0.60

                                      88             0.40

                                      223            1.00

Panel B. "Sell lrecommendation N=101  Observed.Prop.  Test
(2-tailed)                                            Prop.

[P.sub.1]-[P.sub.0]                   65             0.65

                                      36             0.35

[P.sub.2]-[P.sub.0]                   65             0.67

                                      36             0.33

[P.sub.3]-[P.sub.0]                   70             0.64

                                      31             0.36

[P.sub.4]-[P.sub.0]                   71             0.69

                                      30             0.31

[P.sub.5]-[P.sub.0]                   69             0.69

                                      32             0.31

                                      101            1.00

Panel A. "Buy" recommendation N= 223  Asymp. Sig. (2-tailed)
Prop

[P.sub.1]-[P.sub.0]                   0.50         0.503 *

[P.sub.2]-[P.sub.0]                   0.50         0.003 *

[P.sub.3]-[P.sub.0]                   0.50         0.016 *

[P.sub.4]-[P.sub.0]                   0.50         0.032 *

[P.sub.5]-[P.sub.0]                   0.50         0.003 *

Panel B. "Sell lrecommendation N=101  Asymp. Sig. (2-tailed)

[P.sub.1]-[P.sub.0]                   0.50         .000 *


[P.sub.2]-[P.sub.0]

                                      0.50         .000 *
[P.sub.3]-[P.sub.0]

[P.sub.4]-[P.sub.0]                   0.50         .000 *

[P.sub.5]-[P.sub.0]

                                      0.50         .000 *
Note: *Based on Z approximation.
Group 1: firms that experience a lower price
after recommendation date
Group 2: firms that experience a higher price
after recommendation date


Next, we investigate whether Jim Cramer's recommendations have any impact on competitors of the "picks". Based on the SIC code and the size of the firm, we identify 501 companies as the competitors of companies that receive "buy" recommendations, and 238 companies as the competitors of companies that receive "sell" recommendations over the sample period. The results presented in Exhibit 3, Panel A, reveal that on average 29.74% of the competitors experience a higher price one day after the "buy" recommendation; 37.33% of the competitors experience a higher price two days after the "buy" recommendation.

Over 70% of the picks' competitors experience lower prices three, four and five days after the "buy" recommendation date. On the other hand, only about a third of the picks' competitors experience a higher price one through five days after the "sell" recommendation. It appears that investors under-react to news of economically linked firms for both "buy" and "sell" recommendations.
Exhibit 3 Descriptive statistics of price differences for
sample of competitors:"buy" and "sell" recommendations

Panel A: "Buy" recommendation   % of firms that   Std. Dev.
N=501                           experienced
                                positive change
                                in stock price
                                   **

[P.sub.1] -[P.sub.0] *            29.74           0.45757
[P.sub.2] -[P.sub.0]              37.33           0.48415
[P.sub.3] -[P.sub.0]              29.14           0.45487
[P.sub.4] -[P.sub.0]              29.14           0.45487
[P.sub.5] -[P.sub.0]              29.14           0.45487

Panel B: "Sell" recommendation  % of firms that   Std. Dev.
N=238                           experienced
                                positive change
                                in stock price


[P.sub.1] -[P.sub.0] *            35.29           0.47889
[P.sub.2] -[P.sub.0]              31.09           0.46385
[P.sub.3] -[P.sub.0]              38.66           0.48799
[P.sub.4] -[P.sub.0]              35.71           0.48017
[P.sub.5] -[P.sub.0]              31.09           0.46385

Note: *P1-P0 = Price one day after the recommendation date
minus price on the recommendation date. If the price difference
is positive, it is coded as "1"; otherwise, it
is coded "0." **: Number of firms coded 1/total firms


Exhibit 4 presents the results of a non-parametric binominal test of the stock price change for a sample of competitors. According to the results displayed in Panel A:

* 70% of competitors experience a negative price change& one day after "buy" recommendation day

* 63% of competitors experience a lower price' two days after "buy" recommendation day.

* On average, on the third day after a "buy" recommendation, 70% of "buy" picks' competitors experience stock price decreases.

* These results suggest that, on average, the competitors of Jim Cramer's "buy" picks experience lower prices for three consecutive days after the day he recommends his "buy."

On the other hand, it appears that "picks" competitors exhibit less of a negative price trend after "sell" recommendations compared to "buy" recommendations. The initial reaction of over 30% of "picks" competitors had their stock price reversed; starting the third day after "sell" recommendations were issued, they stop falling.

Next, we separate competitors into two groups:

* Group 1 represents companies that experience price decreases after the recommendation day.

* Companies that experience price increases after the recommendation day are placed in Group 2.

According to the non-parametric test results presented in Exhibit 4, Panel A, 70% of companies with "buy" recommendations experience price decreases two, three, four, and five days after the recommendation has been "aired" on TV. Results indicate that, on average, the competitors of Jim Cramer's stocks with "buy" recommendations also experience lower prices two, three, four, and five days after the recommendation date. Only 48% of companies given "buy" recommendation by Jim Cramer exhibit higher prices the day after the recommendation day.

On the other hand, when Cramer told his audience to sell certain stocks, their prices dropped8 the day after he made his "sell" recommendation. We observe that individual investors are more sensitive to "sell" recommendation than to "buy" recommendation. Stocks with "sell" recommendations experience negative price "drift" longer than one day. On average, close to 65% of "sell" recommendation companies experience price decreases the first five days after each recommendation day. At the same time, "buy" recommendations have less of a positive effect on stock prices after recommendation day. The results for competitors are similar to those for sample companies.
Exhibit 4 Non-parametric binomial test stock
price change for competitors: Competitors are
stratified into two groups: firms experiencing
lower stock prices after recommendation date (Group 1)
and firms experiencing higher stock prices (Group 2).
Total companies N = 501

Panel A: "Buy" recommendation N=501   N        Observed Prop  Test.Prop.

[P.sub.1]-[P.sub.0] *                 Group 1  352            0.70

                                      Group 2  149            0.30

[P.sub.2]-[P.sub.0]                   Group 1  314            0.63

                                      Group 2  187            0.37

[P.sub.3]-[P.sub.0]                   Group 1  355            0.71

                                      Group 2  146            0.29

[P.sub.4]-[P.sub.0]                   Group 1  355            0.71

                                      Group 2  146            0.29

[P.sub.5-[P.sub.0]                    Group 1  355            0.71

                                      Group 2  146            0.29

Panel B: "Sell" recommendation N=238  N        Observed Prop  Test.Prop.

[P.sub.1]-[P.sub.0] *                 Group 1  154            0.65

                                      Group 2  84             0.35

[P.sub.2]-[P.sub.0]                   Group 1  164            0.69

                                      Group 2  74             0.31

[P.sub.3]-[P.sub.0]                   Group 1  92             0.61

                                      Group 2  146            0.39

[P.sub.4]-[P.sub.0]                   Group 1  85             0.64

                                      Group 2  153            0.36

[P.sub.5]-[P.sub.0]                   Group 1  74             0.69

                                      Group 2  164            0.31


Panel A: "Buy" recommendation N=501   Asymp. Sig.(2-tailed)

[P.sub.1]-[P.sub.0] *                 0.50                    0.000 *

[P.sub.2]-[P.sub.0]                   0.50                    0.000 *

[P.sub.3]-[P.sub.0]                   0.50                    0.000 *

[P.sub.4]-[P.sub.0]                   0.50                    0.000 *

[P.sub.5-[P.sub.0]                    0.50                    0.000 *

Panel B: "Sell" recommendation N=238  Asymp. Sig..(2-tailed)  0.000 *

[P.sub.1]-[P.sub.0] *                 0.50                    0.000 *

[P.sub.2]-[P.sub.0]                   0.50                    0.000 *

[P.sub.3]-[P.sub.0]                   0.50                    0.001 *

[P.sub.4]-[P.sub.0]                   0.50                    0.000 *

[P.sub.5]-[P.sub.0]                   0.50                    0.000 *

Note: * Based on Z approximation.
Group 1: firms that experience a lower price
after recommendation date
Group 2: firms that experience a higher price
after recommendation date


To investigate whether Jim Cramer's "buy" and "sell" recommendation has any different impact on prices of "picks" and on their competitors' prices one through five days after recommendation day, we compare the mean value of percentage change in stock prices between these two groups. Based on the results presented in Exhibit 5, Panel A, we conclude that the average percentage change in price for sample firms with "buy" recommendation is positive only on the day after the recommendation has been made.

However, starting the second day after the recommendation has been aired, the average percentage change in price for Jim Cramer's "buy picks" tends to reverse, and the prices go down. At the same time, the percentage change in price for "picks" competitors stays negative throughout the five day window. The average percentage change in price for Jim Cramer's "sell picks" and their competitors remains negative throughout the five day window.
Exhibit 5 Percentage changes of stock prices between
price on recommendation dar and prices one through
five days after: Test includes both sample
companies and competitors

Panel A: "Buy" recommendation N=724      N     Mean  Std. Dev

   ([P.sub.1]-[P.sub.0]/[P.subo] *    Group 1  223    0.0026

                                      Group 2  501   -0.0060

  ([P.sub.2]-[P.sub.0])/[P.sub.0]     Group 1  223   -0.0032

                                      Group 2  501   -0.0137

  ([P.sub.3]- [P.sub.0])/[P.sub.0]    Group 1  223   -0.0048

                                      Group 2  501   -0.0048

  ([P.sub.4]-  [P.sub.0])/[P.sub.0]   Group 1  223   -0.0082

                                      Group 2  501   -0.0189

  ([P.sub.5]- [P.sub.0])/[P.sub.0]    Group 1  223   -0.0147

                                      Group 2  501   -0.0246

Panel B: "Sell" recommendation N=339     N     Mean  Std. Dev

  ([P.sub.1]- [P.sub.0])/[P.sub.0]    Group 1  101   -0.0039

                                      Group 2  238   -0.0093

  ([P.sub.2]- [P.sub.0])/[P.sub.0]    Group 1  101   -0.0143

                                      Group 2  238   -0.0178

  ([P.sub.3]- [P.sub.0])/[P.sub.0]    Group 1  101   -0.0189

                                      Group 2  238   -0.0171

  ([P.sub.4]- [P.sub.0])/[P.sub.0]    Group 1  101   -0.0243

                                      Group 2  238   -0.0190

  ([P.sub.5]- [.P.sub.0])/[P.sub.0]   Group 1  101   -0.0409

                                      Group 2  238   -0.0315


Panel A: "Buy" recommendation N=724

   ([P.sub.1]-[P.sub.0]/[P.subo] *    0.03090

                                      0.02504

  ([P.sub.2]-[P.sub.0])/[P.sub.0]     0.04510

                                      0.03443

  ([P.sub.3]- [P.sub.0])/[P.sub.0]    0.05517

                                      0.05517

  ([P.sub.4]-  [P.sub.0])/[P.sub.0]   0.05581

                                      0.04787

  ([P.sub.5]- [P.sub.0])/[P.sub.0]    0.05766

                                      0.05045

Panel B: "Sell" recommendation N=339

  ([P.sub.1]- [P.sub.0])/[P.sub.0]    0.04335

                                      0.02435

  ([P.sub.2]- [P.sub.0])/[P.sub.0]    0.04963

                                      0.04157

  ([P.sub.3]- [P.sub.0])/[P.sub.0]    0.05600

                                      0.04801

  ([P.sub.4]- [P.sub.0])/[P.sub.0]    0.06919

                                      0.05553

  ([P.sub.5]- [.P.sub.0])/[P.sub.0]   0.08406

                                      0.06115

Note: Group 1 = sample companies; Group 2 = competitors
*: Price one day after the recommendation date (P1),
minus price of the recommendation date (PO) divided by
the price of the recommendation date (PO).


We anticipate spikes and valleys in the short-term price change of his picks due to a majority of his audience following his advice.

To determine whether the percentage price change is economically significant between sample and competitors, we perform an independent sample t- test. Exhibit 6 presents the test results. We conclude that for "buy" recommendations, the day-one average percentage change in price for the sample is .859 percent higher than that of the competitors, which is statistically significant. The same conclusion can be drawn for days two through five after the "buy" recommendation day. Based on the empirical findings, it appears that Jim Cramer's recommended stocks outperform the competitors over the sample period. However, both his "picks" and their competitors, on average, experience negative price changes after the recommendation every day, one through five. Results show that both Jim Cramer's buy picks and the competitors in the same industry experienced lower prices after the buy recommendations. However, the competitors experienced even lower stock prices than the picks.
Exhibit 6 Independent sample test of percentage
change in stock price test for sample and competitors
over the five day period

Panel A. "Buy". Recommendation        F    Sign.  Mean **.Difference

([P.sub.1]-[P.sub.0])/[P.sub.0] *

([P.sub.2]-[P.sub.0])/[P.sub.0] *

([P.sub.3]-[P.sub.0])/[P.sub.0] *   5.149  0.024      0.0085905

([P.sub.4]-[P.sub.0])/[P.sub.0] *

([P.sub.5]-[P.sub.0])/[P.sub.0] *

                                    4.463  0.035      0.0010518

                                    5.881  0.016      0.0115085

                                    1.371  0.242      0.0106503

                                    1.647  0.200      0.0099002

Panel B. "Sell". recommendation       F    Sign.  Mean**.Difference

([P.sub.1]-[P.sub.0])/[P.sub.0] *

([P.sub.2]-[P.sub.0])/[P.sub.0] *

([P.sub.3]-[P.sub.0])/[P.sub.0] *   4.297  0.039      -0.00535

([P.sub.4]-[P.sub.0])/[P.sub.0] *

([P.sub.5]-[P.sub.0])/[P.sub.0] *

                                    2.615  0.107      -0.00351

                                    0.736  0.391       0.00185

                                    1.242  0.266       0.00526

                                    4.261  0.032       0.00938

Panel A. "Buy". Recommendation      t-statistic  Sign.

([P.sub.1]-[P.sub.0])/[P.sub.0] *

([P.sub.2]-[P.sub.0])/[P.sub.0] *

([P.sub.3]-[P.sub.0])/[P.sub.0] *      3.956     0.000

([P.sub.4]-[P.sub.0])/[P.sub.0] *

([P.sub.5]-[P.sub.0])/[P.sub.0] *

                                       3.435     0.001

                                       3.039     0.002

                                       2.623     0.009

                                       2.324     0.020

Panel B. "Sell". recommendation     t-statistic  Sign.

([P.sub.1]-[P.sub.0])/[P.sub.0] *

([P.sub.2]-[P.sub.0])/[P.sub.0] *

([P.sub.3]-[P.sub.0])/[P.sub.0] *     -4.522     0.000

([P.sub.4]-[P.sub.0])/[P.sub.0] *

([P.sub.5]-[P.sub.0])/[P.sub.0] *

                                      -7.011     0.000

                                      -6.435     0.000

                                      -6.329     0.000

                                      -9.188     0.000

Note: *: calculated as ([P.sub.1]-[P.sub.0])/[P.sub.0]. **:
The mean percentage change of sample firms - the mean
percentage change of competitors.


For sell recommendations, the results are not as conclusive. For stock prices one day after the recommendation date, the competitors experienced a lower price than the sample and the difference is economically significant. The same conclusion can be drawn for day two stock prices. However, for prices three, four, and five days after recommendation date, the sample experienced lower prices than competitors. The difference is economically significant for price day five at 3.2%. Results indicate that both Jim Cramer's sell pick and the picks' competitors in the same industry experienced lower prices after the recommendations.

The average percentage change in price for sample firms with "buy" recommendations is positive only on the day after the recommendation has been made.

4. Conclusion

In this study, we investigate whether investors can benefit from Jim Cramer's picks in both buy and sell recommendations. To further examine whether the recommendations have any impact on the competitors in the same industry of Cramer's picks, we include the competitors that were selected by using both SIC code and the firm size. 501 competitors were selected for buy recommendations and 238 competitors were selected for sell recommendations. Stock prices one through five days after the recommendations were collected.

The results of our study indicated that the prices of Jim Cramer's buy picks experienced significantly lower prices after the buy recommendations. The competitors of the buy picks experienced even worse stock prices after the recommendations. The price differences between two groups are, on average, statistically significant. According to our results, potential investors will not profit from Jim Cramer's buy picks. However, the results of our study also show that Cramer's picks outperformed the competitors in the same industry.

Regarding sell recommendations, the evidence in this study shows that prices of Cramer's sell picks do experience lower stock prices, and the drops in prices are significantly lower than the price at the recommendation date. The same conclusion can be drawn for the competitors. Competitors also experienced significantly lower stock prices one through five days after the recommendation date. The differences between sample and competitors are inconclusive.

References

Barber, B. and D. Loeffler. 1993. The Dartboard column: Second hand information and price pressure. Journal of Financial and Quantitative Analysis, 28, 273-284.

Barber, B. and T. Odean. 2006. All that glitters: the effect of attention and news on the buying behavior of individual and institutional investors. Review of Financial Studies, 28 (April), 785-818.

Busses, J. and C. Green. 2002. Market efficiency in real time. Journal of Financial Economics.

Cohen, L. and A. Frazzini. 2008. Economic links and predictable returns. Journal of Finance, 63, 4, 1977-2011.

Engleberg J., C. Sasseville, and J. Williams. 2007. Attention and asset prices: the case of Mad Money. Working paper, Kellogg School of Management.

Green, S. and S. Smart. 1999. Liquidity provision and noise trading: evidence from the "Investment Dartboard' Column." Journal of Finance, 54, pp. 1885-1899.

http://www.barrons.com

http://www.madmoneyrecap.com

http://www.onlinetradersforum.com

http://www.thestreet.com

Liang, B. 1999. Price pressure: evidence from the Dartboard Column. Journal of Business, 72, 119-134.

Lim, B. and J. Rosario. 2008. The performance and impact of stock picks mentioned on "Mad Money." Working paper, University of California.

Liu, P., S. D. Smith, and A. A. Syed. 1990. Stock price reactions to the Wall Street Journal's securities recommendations. Journal of Financial and Quantitative Analysis, 25, 399410.

Neumann, J.L. and P. M. Kenny. 2007. Does Mad Money make the market go mad? The Quarterly Review of Economics and Finance, 47, 602-615.

Endnotes

(1) We collect data from first-run episodes.

(2) www.madmoneyrecap.com and www.onlinetradersforum.com

(3) (4) (5) Independent sample t-test: HO: change in price is significantly different from "0."

(6) (7) (8) Independent sample t-test: HO: change in price is significantly different from "0."

Natalya Delcoure, Cameron School of Business, University of St. Thomas, Houston

delcoun@stthom.edu

Joe Ueng, Cameron School of Business, University of St. Thomas, Houston

ueng@stthom.edu
COPYRIGHT 2011 St. John's University, College of Business Administration
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Delcoure, Natalya; Ueng, Joe
Publication:Review of Business
Geographic Code:1USA
Date:Jun 22, 2011
Words:5277
Previous Article:Corporate financial reporting complexity: recommendations for improvement.
Next Article:Adding assurance to the term "organic".
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters