The performance and impact of stock picks mentioned on 'Mad Money'
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Applied Financial Economics, 2010, 20, 1113–1124 The performance and impact of stock picks mentioned on ‘Mad Money’ Bryan Lima,* and Joao Rosariob a Department of Finance, University of Melbourne, Melbourne, VIC 3010, Australia b Department of Economics, University of California, Santa Barbara, CA, USA We analyse both the market reaction and the long-term returns of stock picks mentioned on the Consumer News and Business Channel (CNBC) programme ‘Mad Money’, hosted by former hedge fund manager Jim Cramer. We find that Cramer’s stock-picking style is consistent with a positive-feedback trading strategy, favouring stocks which have out- performed over an interval prior to the pick date. Subsequent to a pick, Cramer’s immediate effect on a stock appears inversely proportional to the corresponding firm’s market capitalization. The returns over a 6-month horizon provide some evidence in favour of Cramer’s stock-picking ability. In particular, his recommendations on small-cap stocks accurately predict the long-run trends. I. Introduction viewership are market forces unto themselves. Were his audience both adequately numerous and dedi- Nearly every business day at 6 pm Eastern Standard cated, Cramer’s picks – at least in the immediate Time, former hedge fund manager Jim Cramer hosts horizon – would become self-fulfilling prophesies: a an hour-long show on the cable financial network glowing recommendation on the show followed by Consumer News and Business Channel (CNBC). a swell of buy orders the next trading day would Typically reaired once again at 11 pm, the pro- provide instant vindication. gramme is billed as a primer on personal investment, To form a complete analysis of his recommenda- with Cramer offering his recommendations and tions, then, one needs to separate his influence from criticisms of companies in an effort to, in his words, his forecasting ability. To this end, we evaluate the ‘make you money’. While historically there has never market response immediately following a broadcast been a scarcity of financial prognosticators willing to of ‘Mad Money’ as well as Cramer’s ability to offer their wisdom to anyone who will listen, few if forecast winners and losers over a 6-month horizon. any can claim the breadth of the domain Cramer We analyse recommendations on the Mad Money commands, with up to 400 000 viewers daily. That his programme along three lines of inquiry. First, we emergence coincided with the ever-increasing ubiq- look for the evidence of positive-feedback trading in uity of personal trading via the internet raises the Cramer’s stock-picking style by examining the trends possibility that where other Wall Street soothsayers of stocks prior to recommendations. Second, we have been rebuffed by market forces, Cramer and his estimate the immediate impact of recommendations *Corresponding author. E-mail: blim@unimelb.edu.au Applied Financial Economics ISSN 0960–3107 print/ISSN 1466–4305 online ß 2010 Taylor & Francis 1113 http://www.informaworld.com DOI: 10.1080/09603101003761887
1114 B. Lim and J. Rosario by the market reaction subsequent to a broadcast. capitalization. By comparison, we document smaller Finally, we test his forecasting ability by calculating excess returns subsequent to recommendations, the long-term (420 trading days) returns associated which may be attributable to either the differing with recommendations. We find that Cramer’s stock- sample selection or the differing measurement of picking style is consistent with a positive-feedback excess returns. Similar to our article, Keasler and trading strategy, favouring stocks which have out- McNeil (2010) document returns associated with performed over an interval prior to the pick date. Cramer’s picks, with a sample period approximately Subsequent to a pick, Cramer’s immediate effect on a 6 months shorter than ours. While both their and our stock appears inversely proportional to the corre- calculations for short-term excess returns are compa- sponding firm’s market capitalization. This result is rable, we find limited evidence of positive excess not unexpected; if the same number of viewers longer-term returns while Keasler and McNeil do responds to each pick, the disturbance on a small- not. This article is further differentiated by our cap stock will be relatively greater than that on a stressing trends in Cramer’s stock picking style as large-cap stock. The returns over a 6-month horizon well as in the overnight component of the 1 day provide some evidence in favour of Cramer’s stock- returns associated with his picks. picking ability. In particular, his recommendations More generally, this article represents an intersec- on small-cap stocks accurately predict the long-run tion of the event study and analyst recommendation trends. literatures. The analysis most closely follows Cramer professes no insider knowledge, frequently Womack’s (1996) examination of both the market exhorting his audience to research firms’ guidance reaction and the long-term returns associated with (earnings estimates) before investing. While the spikes analyst recommendations. The mass-media compo- and valleys in the short-term returns of his picks are nent of this article is in the spirit of research likely due to a substantial portion of his audience documenting the market reaction to stocks men- following his advice, over a sufficiently long horizon tioned in financial news outlets. Barber and Loeffler we would expect the market to adjust given that these (1993), Greene and Scott (1999) and Liang (1999) price changes contain no new information. As such, examine the market response to the Wall Street we interpret the long-term excess returns associated Journal’s ‘Dartboard’ column. Lloyd-Davies and with Cramer’s small-cap stock picks as potential Canes (1978) and Liu et al. (1990) similarly examine evidence of his skill, as opposed to influence. the market response to the Wall Street Journal’s Rudimentary analyses in the nonacademic press ‘Heard on the Street’ column. Busse and Green typically find little favourable evidence of Cramer’s (2002) document intraday trading response to ability. We postulate here that such results are driven CNBC’s ‘Morning Call’ and ‘Midday call’ segments, by the fact that nearly 75% of Cramers’ picks while Barber and Odean (2006) examine the effects of represent either large-cap stocks or responses to high levels of news coverage on a given stock’s price. callers’ questions or both. In general, size-adjusted excess returns for large-cap stocks are likely to be relatively small, in which case it may not be possible for Cramer to generate significant excess returns II. Data within this category. By the same token, the market response to caller picks is small versus the response Our sample consists of episodes broadcast between to Cramer’s in-depth recommendations. More than 28 June 2005 and 22 December 2006. The data was likely, the amount of information (that is perceived to taken from official recaps posted at thestreet.com, a be) conveyed to viewers during the caller segments is financial website founded by Cramer. Our tran- relatively low. Taken together, the huge numbers of scribed results were spot-checked against independent large-cap and caller picks are likely weighting the recap websites,1 though none had picks for shows calculated excess returns towards zero. earlier than November 2005. We restrict tickers to This article is similar in spirit to two contempora- those listed on the New York Stock Exchange neous papers. Engleberg, Sasseville and Williams (NYSE), National Association of Securities Dealers (ESW, 2007) examine first-time picks on the Mad Automated Quotations (NASDAQ) or American Money programme. Concentrating on a 6-month Express (AMEX) exchanges. window with 391 picks, they find significant next-day Picks are identified as either ‘Buy’ or ‘Sell’. While returns for first-time positive recommendations by this simple binary categorization effectively elimi- Cramer, returns which increase inversely with market nates distinguishing between weak and strong 1 www.madmoneyrecap.com and www.onlinetradersforum.com.
Mad Money 1115 recommendations (positive or negative), it also We group stocks into three bins: ‘small cap’ for deciles removes the subjectivity that would be involved in 1–5, ‘mid cap’ for deciles 6–8; and ‘large cap’ for such a transcription process. In this sense, a ‘Buy’ deciles 9 and 10. We stress that these classifications are may not represent the actual word ‘buy’ being spoken relative and therefore do not correspond to the by Cramer but simply an endorsement, however standard definitions in the practitioner’s lexicon. measured. Our primary sample consists of picks exclusively We further separate picks by the party who from first-run episodes. Table 1 summarizes the data. initiated the ticker in question: caller or noncaller. The 10 589 picks represent 2074 distinct firms from 560 During such segments as ‘Lightning Round’, ‘Sudden unique Standard Industrial Classification (SIC) codes. Death’ and ‘Are You Diversified?’, viewers call into (Remarkably, over the sample period, the entire the programme and ask Cramer for his opinion on Center for Research in Security Prices (CRSP) particular stocks. Caller picks are typically discussed database lists 8012 distinct firms from 922 unique only cursorily, and it seems unlikely that on aver- SIC codes.) Examining Table 1, we observe several age a 5-second recommendation during the trends: ‘Lightning Round’ will convey the same information as a 5-minute recommendation during the opening . Among noncaller picks, Buys outnumber Sells by segment. more than 5 to 1. This ratio is roughly fixed In terms of volume, the vast majority of the tickers across each market capitalization group. By and mentioned on the programme represent questions large, the programme is a forum for Cramer to from callers. Cramer ostensibly has no previous highlight companies he prefers as opposed to knowledge that these tickers will be inquired about those he dislikes. There may be any number of (though they are often follow-ups to his previous reasons to explain this tendency, with a likely picks), and as such, his responses may not fairly culprit being his audience’s inability or unwill- represent his stock-picking ability. ingness to sell negatively recommended stocks. The majority of a broadcast’s running time consists In Section ‘Overnight returns’, we document of Cramer discussing what we call noncaller picks. only modest market reaction to a Sell pick. It These picks represent firms that Cramer premeditat- may be the case that Cramer simply tailors his edly chooses to mention on the programme. Unlike show to match the investing habits of his caller picks, he has complete discretion over the firms audience. in this category. . Among caller picks, the probability of Cramer Regular viewers of ‘Mad Money’ will observe the recommending a Buy increases with market absence of other potential categorizations, like ‘Pick capitalization. For caller tickers from deciles 1 of the Week’ or ‘Sell Block’. While such categories to 5, the Buy-to-Sell ratio is less than 1 to 2, appear in our database, given the evolving nature of while for tickers in deciles 9 and 10, the ratio the programme we have chosen only the most general switches to about 2.3 to 1. categories in which to sort. . Approximately two-thirds of all picks represent For each trading day, we rank each stock of the firms from the top 20% market capitalization. combined NYSE, NASDAQ and AMEX databases to This ratio holds for both noncaller and determine its market capitalization decile on that date. caller picks. Table 1. Summary statistics Market cap decile All 1–5 6–8 9 and 10 Distinct tickers Distinct SIC codes Noncaller picks 2701 135 692 1874 1055 382 Buys 2256 110 553 1593 929 348 Sells 445 25 139 281 322 193 Caller picks 7888 306 2172 5410 1874 527 Buys 4916 93 1033 3790 1161 407 Sells 2972 213 1139 1620 1343 421 Full sample 10 589 441 2864 7284 2074 560 Average market cap $23 166 $227 $880 $33 317 – – Note: Market capitalization in millions.
1116 B. Lim and J. Rosario III. Pre- and Post-pick Returns pick date (note that a firm’s SIC code is not necessarily constant over time) for those with a We calculate the raw, size-adjusted and industry- minimum of five other firms with the same SIC. We adjusted excess returns for the stocks mentioned on calculate industry-adjusted excess returns over period the programme. As ‘Mad Money’ is broadcast after as the size-adjusted excess return for firm i minus markets close, we must be precise in our holding- the average size-adjusted excess return for the J other period definitions. Given a pick date T and a holding firms with the same SIC code as firm i period before or after T, we calculate the following time-horizon (net) returns: 1 X J ERSIC size i, ¼ ERi, ERsize j, J j2SICðiÞ . Period before pick date ¼ (Closing price on j6¼i T /Closing price on T ) 1 . Period after pick date: (Closing price on T þ / Caller picks Closing price on T ) 1 As callers do not know ex ante whether Cramer will All returns are calculated cum dividend, with the positively or negatively recommend their picks, we exception of the overnight returns discussed later in calculate the pre-pick returns for all caller picks, this section. regardless of Cramer’s eventual recommendation, in There is no distinction between calculating returns order to capture the information set as of the time of for Buys and for Sells; they are computed identically. filming. The average size-adjusted excess returns over Specifically, returns are calculated per stock, so a the 60 and 20 trading days immediately prior to the positive return on a Buy or a Sell simply means that pick date are 6.78% (t ¼ 23.4) and 2.34% (t ¼ 13.7), the stock itself has outperformed its benchmark over respectively. That these returns are positive and the specified horizon. In terms of post-pick perfor- significant suggests that Cramer’s audience – at mance, positive (excess) returns for Buys and negative least, those who call in – is generally interested in (excess) returns for Sells would ostensibly be desirable his opinion on ‘hot’ stocks, those that have out- from a viewer’s perspective. performed over the previous month or quarter. Viewers presumably call in to gauge his opinion on Raw returns whether the stock has peaked or not. We calculate the return to stock i over period as Figure 1 graphs cumulative average size-adjusted excess returns for caller picks over an interval of Y 1 þ Ri, ¼ ð1 þ ri,t Þ 20 days prior and subsequent to the pick date, denoted t by day 0. While both buys and sells exhibit increasing where ri,t is the return from holding stock i from the returns over the interval up to day 0, we observe that market close on date t 1 to the market close on date t. Cramer recommends buying stronger outperformers and selling weaker ones. More precisely, the average Size-adjusted excess returns 20 trading-day pre-pick size-adjusted excess return on We compare pick returns to the CRSP value- weighted index for each pick’s market capitalization Caller picks 0.04 decile, ranked against the aggregated AMEX, Cumulative excess return NASDAQ and NYSE firms. The size-adjusted 0.04 excess return over period is calculated as the 0.03 return for firm i minus the return for the CRSP value- 0.03 Buys weighted index for the corresponding decile 0.02 Sells Y Y 0.02 1 þ ERsize i, ¼ ð1 þ ri,t Þ ð1 þ rdecileðiÞ, t Þ 0.01 t t 0.01 0.00 –20 –15 –10 –5 0 5 10 15 20 Industry-adjusted excess returns Trading days relative to pick date We compare pick returns versus averages of other Fig. 1. Cumulative size-adjusted excess returns for caller firms in the same industry. Following Womack picks (1996), we construct equally-weighted portfolios of Note: The horizontal lines indicate the cumulative excess firms with the same SIC code as the pick firm on the return as of the market close on the date of the pick.
Mad Money 1117 a caller ticker which Cramer recommends buying is about stocks which are outperforming, with the 3.47%, while the equivalent for a sell is only 1.34%, as level of outperformance roughly negatively related shown in Table 2. to the market capitalization. Looking at the post-pick period, we observe small, For small-cap stocks, the pre-pick cumulative positive but significant excess returns for caller buys excess returns of buys and sells are nearly identical, over the next possible trading day, gradually decaying suggesting Cramer could not be applying any simple towards zero over the following month. Similarly, for technical analysis rule to make his recommendations caller sells, we observe small, negative but significant within this group. More interestingly, the post-pick excess returns over the next possible trading day, returns for small-cap caller picks suggest that his remaining relatively flat over the following month. recommendations for this group are correct: Cramer The size-adjusted excess returns over a 20 trading-day is on average correctly identifying whether or not the horizon for buys and sells are not significantly stock in question has reached its short-term peak. different, a fact which may have more to do with Small-cap caller sells exhibit a downward drift after both the breadth of the selections and the selection the pick date, while the corresponding buys peak 7 process of Cramer’s audience than with Cramer trading days after the pick date before drifting himself. downward. Moreover, of the small cap stocks he is asked by callers, his sells underperform while his buys Conditional on market capitalization. We decom- outperform over the subsequent 20 trading days. pose the pre- and post-pick returns by market The results for mid-cap caller picks are consider- capitalization. Figure 2 graphs the results. Within ably less dramatic. A recommendation on the show each market capitalization group, callers inquire elicits a small next-day excess return with the proper Table 2. Caller pick returns Size-adjusted Industry-adjusted Raw return Excess return Excess return Sells Buys Sells Buys Sells Buys All caller picks (n ¼ 2971) (n ¼ 4916) (n ¼ 2971) (n ¼ 4916) (n ¼ 2399) (n ¼ 3917) 20-Day period before pick 2.76 4.36 1.35 2.95 1.28 2.45 (7.58) (24.62) (3.79) (17.37) (3.18) (12.86) 1-Day period after pick 0.3 0.36 0.38 0.3 0.38 0.27 (5.50) (10.50) (7.13) (9.20) (6.79) (7.52) 20-Day period after pick 0.64 0.9 0.48 0.1 0.9 0.23 (2.85) (6.35) (2.27) (0.77) (3.97) (1.60) Deciles 1–5 (n ¼ 213) (n ¼ 93) (n ¼ 213) (n ¼ 93) (n ¼ 162) (n ¼ 74) 20-Day period before pick 11.52 10.79 9.52 8.83 9.89 8.16 (5.46) (3.13) (4.59) (2.60) (4.22) (2.00) 1-Day period after pick 1.00 1.97 1.11 1.80 0.98 1.62 (3.55) (4.99) (4.05) (4.62) (3.10) (3.75) 20-Day period after pick 1.64 2.57 2.68 2.10 3.85 2.88 (1.35) (1.49) (2.33) (1.25) (3.05) (1.54) Deciles 6–8 (n ¼ 1139) (n ¼ 1033) (n ¼ 1139) (n ¼ 1033) (n ¼ 892) (n ¼ 782) 20-Day period before pick 4.01 6.79 2.37 5.11 2.04 5.10 (5.22) (12.34) (3.15) (9.61) (2.24) (8.07) 1-Day period after pick 0.45 0.79 0.53 0.72 0.53 0.59 (4.28) (7.65) (5.30) (7.34) (4.73) (5.31) 20-Day period after pick 0.11 0.46 1.03 0.51 1.57 0.96 (0.27) (1.12) (2.59) (1.32) (3.52) (2.14) Deciles 9 and 10 (n ¼ 1620) (n ¼ 3790) (n ¼ 1620) (n ¼ 3790) (n ¼ 1346) (n ¼ 3061) 20-Day period before pick 0.73 3.55 0.46 2.22 0.27 1.65 (2.75) (23.50) (1.80) (15.59) (1.03) (10.78) 1-Day period after pick 0.11 0.21 0.17 0.15 0.21 0.15 (1.86) (6.29) (3.07) (4.84) (3.79) (4.51) 20-Day period after pick 1.29 0.98 0.19 0.05 0.11 0.13 (5.47) (6.99) (0.86) (0.35) (0.48) (0.89) Note: t-Statistics are given in parentheses.
1118 B. Lim and J. Rosario 0.18 Caller – Sells Noncaller picks 0.06 0.16 Cumulative excess return Deciles 1 –5 Cumulative excess return 0.05 0.14 Deciles 6 –8 0.04 0.12 Deciles 9 and 10 0.03 0.10 0.02 0.08 0.01 Buys 0.06 Sells 0.04 0.00 0.02 –0.01 0.00 –0.02 –0.02 –0.03 –20 –15 –10 –5 0 5 10 15 20 –20 –15 –10 –5 0 5 10 15 20 Days relative to pick date Trading days relative to pick date Caller – Buys Fig. 3. Cumulative size-adjusted excess returns for non- 0.18 0.16 caller picks Deciles 1 –5 Cumulative excess return Note: The horizontal lines indicate the cumulative excess 0.14 Deciles 6 –8 return as of the market close on the date of the pick. 0.12 Deciles 9 and 10 0.10 0.08 0.06 Immediately after a recommendation, we observe 0.04 significant next-day size-adjusted excess returns: 0.02 0.88% for sells and 1.56% for buys. Over the next 0.00 month, however, these excess returns drift towards –0.02 zero, though for buys, the excess returns remain –20 –15 –10 –5 0 5 10 15 20 Days relative to pick date positive. Fig. 2. Cumulative size-adjusted excess returns for caller Conditional on market capitalization. Figure 4 picks, conditional on market capitalization graphs the cumulative size-adjusted excess returns Note: The horizontal lines indicate the cumulative excess for noncaller picks, decomposed by market capital- return as of the market close on the date of the pick. ization. As evidenced by the pre-pick excess returns, Cramer tends to favour outperformers and to reject corresponding sign (positive for buys and negative for underperformers within each market capitalization sells); subsequent to that first trading day, the post- group. pick drift for sells is flat and for buys slopes Following a broadcast, the next-day excess returns downward. After 20 trading days, the post-pick for small-cap noncaller picks are relatively large in excess returns for both mid-cap buys and sells are magnitude, particularly for buys. The excess returns not significantly different from each other. for buys exhibit a slight upward drift after the first Large cap stocks appear generally undisturbed by trading day, while for sells they exhibit a slight Cramer’s recommendations to callers’ questions. downward drift over the first 2 weeks (10 trading While he favours large cap outperformers to under- days) before sharply increasing thereafter. If pur- performers, the post-pick drift for both large-cap chasing the stock at the closing price of the next caller buys and sells is essentially flat. trading day after a broadcast, however, one earns significantly higher 1-month excess returns purchas- ing small-cap stocks Cramer recommends selling than Noncaller picks those he recommends buying. Excess returns for mid-cap noncaller buys increase As observed in Fig. 3, among noncaller picks, the sharply the next trading day, then drift downward stocks that Cramer recommends selling have under- towards zero over the subsequent month. For sells, performed over the previous month while those he the next-day excess returns appear to be part of a recommends buying have outperformed over the general downward drift that begins 8 trading days same horizon. Viewed alongside his pattern of prior to the pick date and ends 8 trading days recommending the highest outperformers among afterward. caller-requested stocks, these tendencies suggest that Echoing the observed trends for large-cap caller Cramer favours a positive feedback strategy: buying picks, noncaller large-cap stocks are largely unaf- (relative) winners and selling (relative) losers fected by a positive or negative recommendation by (Table 3). Cramer. Sells exhibit a temporary dip in excess
Mad Money 1119 Table 3. Noncaller pick returns Size-adjusted Industry-adjusted Raw return Excess return Excess return Sells Buys Sells Buys Sells Buys All noncaller picks (n ¼ 445) (n ¼ 2260) (n ¼ 445) (n ¼ 2260) (n ¼ 337) (n ¼ 1742) 20-Day period before pick 0.09 4.45 1.22 3.24 1.46 2.63 (0.13) (17.91) (1.88) (13.76) (1.95) (9.46) 1-Day period after pick 0.81 1.60 0.88 1.56 0.82 1.58 (3.91) (17.24) (4.29) (17.26) (4.87) (14.54) 20-Day period after pick 0.88 1.92 0.43 0.85 1.22 0.94 (1.37) (8.85) (0.70) (4.17) (1.72) (3.99) Deciles 1–5 (n ¼ 25) (n ¼ 110) (n ¼ 25) (n ¼ 110) (n ¼ 20) (n ¼ 88) 20-Day period before pick 4.78 5.31 4.58 3.37 6.22 0.72 (1.16) (3.34) (1.19) (2.24) (1.19) (0.41) 1-Day period after pick 4.39 7.31 4.54 7.20 3.00 7.58 (1.83) (7.22) (1.89) (7.16) (2.28) (6.23) 20-Day period after pick 2.57 8.53 1.03 7.42 5.14 7.54 (0.41) (5.66) (0.17) (5.11) (0.77) (4.32) Deciles 6–8 (n ¼ 139) (n ¼ 553) (n ¼ 139) (n ¼ 553) (n ¼ 95) (n ¼ 414) 20-Day period before pick 0.55 6.68 1.85 5.37 1.73 5.19 (0.36) (9.44) (1.27) (7.92) (0.94) (6.22) 1-Day period after pick 0.70 2.82 0.92 2.85 1.16 2.87 (2.26) (13.36) (3.09) (13.89) (3.27) (11.69) 20-Day period after pick 0.05 1.65 1.47 0.68 4.01 0.72 (0.04) (2.94) (1.19) (1.29) (2.51) (1.16) Deciles 9 and 10 (n ¼ 281) (n ¼ 1593) (n ¼ 281) (n ¼ 1593) (n ¼ 222) (n ¼ 1238) 20-Day period before pick 1.08 3.63 0.39 2.50 0.91 1.93 (1.67) (15.99) (0.63) (11.76) (1.34) (7.84) 1-Day period after pick 0.55 0.78 0.53 0.73 0.48 0.73 (2.86) (11.18) (2.84) (10.94) (2.94) (9.22) 20-Day period after pick 1.18 1.58 0.05 0.47 0.63 0.54 (2.13) (7.49) (0.10) (2.40) (1.10) (2.44) Note: t-Statistics are given in parentheses. returns, while buys exhibit a small but persistent For two subsamples – caller and noncaller buys increase. from deciles 6 through 8 – the overnight raw returns are greater than the 1 day raw returns. For such picks, prices on average fall from opening to closing Overnight returns on the next possible trading day after a broadcast, though the relative magnitude of the drop is small. Given the daily nature of broadcasts, we separate the overnight component of the 1-day raw returns to examine the immediate reaction to picks on the show. Given a broadcast date t, we define the overnight return as IV. Abnormal Volume Opening price on t þ 1 rovernight ¼ 1 In order to measure the impact on the market of a Closing price on t recommendation, we use a slightly modified version As shown in Table 4, the overnight change in price of Womack’s (1996) measure for Abnormal Volume accounts for virtually all of the 1 day returns, (AV): the ratio of volume on date t to the average regardless of the subsample. The difference in the volume for the 60 previous and subsequent next-day opening price from the previous-day close trading days. can be accounted for by either after-hours trading or Vi,t an accumulation of orders in the specialists’ books AVi,t ¼ Pt1 P ð t0 ¼t60 Vi,t0 þ tþ60t0 ¼tþ1 Vi,t0 Þ=120 before markets open.
1120 B. Lim and J. Rosario Noncaller – Sells Cumulative excess return 0.15 Table 4. Decomposed 1-day raw returns Deciles 1–5 0.10 Buys Sells Deciles 6–8 Deciles 9 and 10 0.05 1 Days Overnight 1 Day Overnight Noncaller: 1–5 7.31 7.31 4.39 3.02 0.00 December (7.22) (8.56) (1.83) (1.41) Noncaller: 6–8 2.82 3.27 0.70 0.48 –0.05 December (13.36) (18.64) (2.26) (2.46) Noncaller: 9 and 0.78 0.67 0.55 0.46 –0.10 10 December (11.20) (9.36) (2.86) (3.40) –20 –15 –10 –5 0 5 10 15 20 All noncaller picks 1.60 1.63 0.81 0.61 Days relative to pick date (17.25) (18.96) (3.91) (3.80) Noncaller – Buys Caller: 1–5 1.97 1.39 1.00 0.08 0.15 December (4.99) (6.32) (3.55) (0.50) Cumulative excess return Caller: 6–8 0.79 0.81 0.45 0.29 0.10 December (7.65) (12.18) (4.28) (1.57) Caller: 9 and 10 0.21 0.17 0.11 0.11 0.05 December (6.29) (5.29) (1.86) (3.04) All caller picks 0.36 0.32 0.30 0.18 0.00 (10.50) (11.37) (5.49) (2.38) Deciles 1–5 Deciles 6–8 –0.05 Notes: t-Statistics are given in parentheses. Returns are Deciles 9 and 10 calculated based on purchasing the stock at the closing –0.10 price of the pick date. ‘1 Day’ is the return associated with –20 –15 –10 –5 0 5 10 15 20 selling the stock at the closing price of the next possible Days relative to pick date trading day. ‘Overnight’ is the return associated with selling the stock at the opening price of the next possible Fig. 4. Cumulative size-adjusted excess returns for non- trading day. caller picks, conditional on market capitalization Note: The horizontal lines indicate the cumulative excess return as of the market close on the date of the pick. investors, Cramer’s viewers may lack the means or even desire to short sell the stocks Cramer Figure 5 graphs AV, conditional on the pick type. views unfavourably. Certainly, most retail Day 0 refers to the date of the broadcast, with internet brokers do not readily provide avenues markets having already closed before Cramer makes for clients to short sell. Moreover, some inves- his recommendations. Day 1 represents the first day tors may be wary of covering a short position, a viewer can trade on the open market after a given the theoretically infinite potential for broadcast. losses. For both caller and noncaller sells across all market (3) Cramer’s effect is inversely proportional to the cap groups, AV peaks on day 0, the trading day market capitalization of the stock. This result which has just concluded as ‘Mad Money’ is broad- may be considered fairly intuitive, as the cast. Equivalently, for buys, AV peaks on day 1, the average trading volume typically increases next possible trading day after the pick. In omitted with a stock’s market capitalization. It may results, we find similar trends obtain for abnormal be the case that on average the same number of (number of ) trades. viewers respond to a recommendation regard- Defining Cramer’s effect on markets as the com- less of the stock’s market cap; given that, one bination of AV and excess returns on day 1, we find would expect the relative impact on smaller several congruent results: cap stocks to be greater than that on larger cap stocks. Another possibility is that viewers (1) Cramer’s effect is higher for noncaller picks than simply respond less to recommendations on caller picks. Confirming our earlier intuition, large cap stocks, about which they may viewers respond less to the recommendations already have considerable information. Cramer makes in response to callers’ stock An endorsement of a relatively unknown questions than to those which come from his stock like American Shared Hospital pre-planned segments. Services (AMS) may spur more viewers to (2) Cramer’s effect higher for buys than sells. If his place orders than an endorsement of, say, audience is composed primarily of personal Proctor & Gamble (PG).
Mad Money 1121 AV – caller Sells AV – caller Buys 7.5 7.5 7 7 6.5 6.5 6 6 5.5 5.5 5 5 4.5 4.5 AV AV 4 4 Deciles 1–5 3.5 Deciles 1–5 3.5 3 Deciles 6–8 Deciles 6–8 3 2.5 2.5 Deciles 9 and 10 2 Deciles 9 and 10 2 1.5 1.5 1 1 0.5 0.5 –10 –5 0 5 10 –10 –5 0 5 10 Days relative to pick date Days relative to pick date 7.5 AV – noncaller Sells AV – noncaller Buys 7.5 7 7 6.5 6.5 6 6 5.5 5.5 5 5 4.5 4.5 AV 4 AV Deciles 1–5 Deciles 1–5 4 3.5 3.5 Deciles 6–8 3 Deciles 6–8 3 Deciles 9 and 10 2.5 Deciles 9 and 10 2.5 2 2 1.5 1.5 1 1 0.5 0.5 –10 –5 0 5 10 –10 –5 0 5 10 Days relative to pick date Days relative to pick date Fig. 5. AV: Day 0 refers to the date of the pick, with markets closing before the show airs Significance testing likely understated, given that the data only capture trades occurring when markets are open. Presumably, In order to test the significance of the AV statistics, there are a number of trades occurring in the after- we construct empirical distributions across the aggre- hours trading during and subsequent to a broadcast. gated NYSE, NASDAQ and AMEX exchanges. We compute AV statistics for every stock on every trading day between 1 January 2005 and 31 December 2006. These statistics are separated by decile: for each decile, we have an V. Long-Term Returns empirical Cumulative Distribution Function (CDF) to compare. We define In order to estimate Cramer’s forecasting ability, we calculate the long-term returns associated with . d(i) ¼ decile of stock i his recommendations. Figure 6 plots the results, . Fd(i) ¼ empirical CDF of AV for decile d(i) separated by pick type and market capitalization group. Detailed returns are presented in the We calculate critical values xd, such that Appendix. Fd (xd,) ¼ 1 . Details of the distributions are As was the case with shorter horizons, Cramer’s presented in the Appendix. small-cap picks are the most accurate over longer For each pick we test whether the AV on the next holding periods: subsequent to the pick date, small- possible trading day (day 1) exceeds the critical values cap sells trend downward while buys trend upward for the pick’s decile. Table 5 presents the results. over the next 6 months. There is considerably Formalizing the results from the last subsection, the volatility in these returns, owing partly to the percentage of picks significant at generally (1) is individual volatilities of the underlying stocks higher for noncaller picks than caller picks, (2) is and partly to the relatively small number of higher for buys than sells and (3) decreases with small-cap picks. respect to the market cap grouping. As noted in Mid-cap picks all trend downward after the pick ESW (2007), the calculated values for day 1 AV are date, regardless of the pick type or recommendation.
1122 B. Lim and J. Rosario Table 5. AV significance testing Sells Buys Small cap Mid cap Large cap All Small cap Mid cap Large cap All Caller Picks 213 1139 1620 2972 93 1033 3790 4916 Significant at 5% 21 91 123 235 12 117 260 389 (10%) (8%) (8%) (8%) (13%) (7%) (7%) (8%) Significant at 1% 5 23 35 63 4 30 45 79 (2%) (2%) (2%) (2%) (4%) (3%) (1%) (2%) Noncaller Picks 25 128 275 428 101 518 1584 2203 Significant at 5% 6 20 46 72 60 227 221 508 (24%) (16%) (17%) (17%) (59%) (44%) (14%) (23%) Significant at 1% 3 8 20 31 42 113 50 205 (12%) (1%) (1%) (1%) (42%) (22%) (1%) (4%) Note: Percentage of picks’ AV significant at in parentheses. Noncaller Sells Noncaller Buys 0.15 0.15 Cumulative excess return Cumulative excess return 0.1 Deciles 1–5 0.1 Deciles 6–8 0.05 Deciles 9 and 10 0.05 0 0 Deciles 1–5 Deciles 6–8 –0.05 –0.05 Deciles 9 and 10 –0.1 –0.1 –20 0 20 40 60 80 100 120 –20 0 20 40 60 80 100 120 Trading days relative to pick date Trading days relative to pick date Caller Sells Caller Buys 0.15 0.15 Cumulative excess return Deciles 1–5 Cumulative excess return 0.1 Deciles 6–8 0.1 Deciles 9 and 10 Deciles 1–5 Deciles 6–8 0.05 0.05 Deciles 9 and 10 0 0 –0.05 –0.05 –20 0 20 40 60 80 100 120 –20 0 20 40 60 80 100 120 Trading days relative to pick date Trading days relative to pick date Fig. 6. Long-term cumulative size-adjusted excess returns, conditional on market capitalization Note: The horizontal lines indicate the cumulative excess return as of the market close on the date of the pick. That is, both Cramer and his callers are selecting primarily to the sheer number of picks. With several mid-cap stocks which underperform the index returns thousand large-cap pick-returns represented (and no for their corresponding deciles. transaction fees considered), the average of these Large-cap stocks appear largely flat after either a returns would be expected to remain close to the Buy or Sell recommendation. We attribute this trend corresponding index returns.
Mad Money 1123 VI. Conclusion Unlike most professional stock pickers, however, Cramer does not have the luxury of selectivity. Measuring Cramer’s impact on equity markets by the Whereas his financial sector counterpart may be combination of next-day returns and AV, we empha- content to recommend a portfolio with a relatively size three general trends: his impact is greater for small number of stocks, Cramer must consistently noncaller picks than caller picks, his impact is greater generate new picks in order to remain relevant. Even for smaller cap stocks and his impact is greater for if it were the case that Cramer’s preferred portfolio buys than sells. consisted of, say, 50 stocks, the laws of television Evidence of Cramer’s forecasting ability is favour- dictate that he must advocate hundreds more on his able. Mid- and large-cap post-pick excess returns are programme. In this sense, one must handicap generally of the correct sign, though the magnitude Cramer’s forecasting ability with respect to similar of these returns is relatively small. Where Cramer studies of mutual- and hedge-fund managers. displays the most ability is with small-cap stocks, in both his caller and noncaller picks. Somewhat curi- ously, conditional on purchasing a stock at the next- day opening price – which, given the negligible References differences between overnight and next-day returns, Barber, B. and Loeffler, D. (1993) The dartboard column: is essentially identical to the next-day closing price – second hand information and price pressure, Journal of the highest returns are associated with long positions Financial and Quantitative Analysis, 28, 273–84. in small-cap noncaller sells. Barber, B. M. and Odean, T. (2006) All that glitters: the Long-term post-pick excess returns on noncaller effect of attention and news on the buying behavior picks have the proper sign and are significant but are of individual and institutional investors, Review of Financial Studies, 21, 785–818. on average only about 1% above or below their Busse, J. A. and Green, T. C. (2002) Market efficiency in benchmarks. The long-term post-pick excess returns real time, Journal of Financial Economics, 65, 415–37. for caller picks are all negative, which, coupled with Engleberg, J., Sasseville, C. and Williams, J. (ESW) (2007) the positive pre-pick excess returns, suggest that Attention and asset prices: the case of Mad Money, callers are inquiring about not only ‘hot’ stocks but Working Paper, Kellogg School of Management. Greene, J. and Scott, S. (1999) Liquidity provision and also overvalued ones. noise trading: evidence from the ‘investment dart- board’ column, Journal of Finance, 54, 1885–99. Caveats Keasler, T. R. and McNeil, C. R. (2010) Mad Money stock recommendations: market reaction and performance, Any conclusions to be derived from the preceding Journal of Economics and Finance, 34, 1–22. analysis of Cramer’s stock picking ability must be Liang, B. (1999) Price pressure: evidence from the considered incomplete, as Cramer’s recommenda- Dartboard column, Journal of Business, 72, 119–34. Liu, P., Smith, S. D. and Syed, A. A. (1990) Stock tions are generally too nuanced to be captured by price reactions to the Wall Street Journal’s securities the simple strategy implied by our examination. Like recommendations, Journal of Financial and most professional stock pickers, Cramer consistently Quantitative Analysis, 25, 399–410. advocates an active trading style, sometimes indicat- Lloyd-Davies, P. and Canes, M. (1978) Stock prices and the ing price targets, and the simple buy-and-hold strat- publication of second-hand information, Journal of Business, 51, 43–56. egies in this article are unable to capture the trading Womack, K. (1996) Do brokerage analysts’ recommenda- strategies presented on his programme with any tions have investment value?, Journal of Finance, 51, precision. 137–67.
1124 B. Lim and J. Rosario Appendix Table A1. AV empirical CDF Decile Observations Xdecile,5% Xdecile,1% 1 266 096 3.3741 9.0915 2 288 810 3.1520 7.7375 3 290 659 2.9491 7.0700 4 289 928 2.6665 6.1099 5 290 896 2.3717 5.0037 6 291 513 2.2740 4.5574 7 293 178 2.1729 4.2790 8 293 324 2.0839 3.9365 9 294 299 2.0371 3.8025 10 296 863 1.8580 3.1687 Notes: For each decile, we calculate the empirical CDF for the distribution of abnormal volume observations. Xdecile, is a critical value such that Fdecile(Xdecile,) ¼ 1 .
Copyright of Applied Financial Economics is the property of Routledge and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
You can also read