|Cramer's Star Outshines His Stock Picks |
By BILL ALPERT
MONDAY, FEBRUARY 9, 2009
Cramer's Star Outshines His Stock Picks
By BILL ALPERT online.barrons.com
Stock picks featured on Mad Money don't live up to the host's hype.
JIM CRAMER'S CELEBRITY IS BIGGER THAN EVER. As financial markets came apart in October, more than 600,000 viewers turned to his Mad Money show -- the biggest crowd since the Nielsen Company started tracking the CNBC series. He is giving advice to huge audiences on NBC's Today Show and getting awestruck coverage in Esquire magazine.
And why not? An earthquake has hit Wall Street, and the 53-year old broadcaster has spent more time there than most any TV journalist. The guy is a hardworking genius with a word of advice for everyone...many words of advice, actually. He dispenses thousands of Buy/Sell recommendations a year and has declared that those stock picks will help you get rich.
Brad Trent for Barron's
In 2007, when we questioned Cramer's performance, he told viewers we were know-nothings and assured them his Mad Money picks had "killed" the S&P 500.
The only regrettable thing about any of this is that CNBC and Cramer won't meaningfully discuss how his advice pans out.
Cramer's recommendations underperform the market by most measures. From May to December of last year, for example, the market lost about 30%. Heeding Cramer's Buys and Sells would have added another five percentage points to that loss, according to our latest tally.
To his credit, Cramer's Sells "made money" by outperforming the market on the downside by as much as five percentage points (depending on the holding period and benchmark). His Buys, however, lost up to 10 percentage points more than the market.
These batting averages represent his stock-picking over a stretch of time, but Cramer is wildly inconsistent, and the performance of individual picks varies widely. So widely, in fact, that it is impossible to know with confidence that any sample of Cramer's recommendations will enable you to outperform the market.
These facts don't mean that viewers should avoid his informative and entertaining show -- they should just be wary of his stock picks.
OTHER CAREFUL, HONEST EXAMINATIONS of the CNBC star showed the same underperformance -- including several independent studies by finance researchers, and a 2007 review by Barron's that found the only way to reliably profit from Cramer's stock picks was to short them (see "Shorting Cramer," Aug. 20, 2007).
That seems to be what smart traders have done, from the evidence of options-market activity examined by a finance professor, who found that betting against Mad Money's Buy recommendations can yield 25% in a month.
The recent performance of Mad Money's stocks resembles past periods in another striking way. Our research reveals that the stocks Cramer picks as Buys have been rising versus the market for several days in advance of his show, while his Sells have been falling. This doesn't prove there is a leak in the tight security surrounding CNBC's show. It could merely mean that Cramer and his staff are heavy-footed in their research. Or it could mean that his stocks are primarily momentum plays. That is the network's explanation. "Jim likes to recommend 'what is working'," said CNBC communications vice president Brian Steel in a written response Friday. "So it is no surprise there would be movement in these stocks prior to Jim mentioning them."
In any event, these pre-show moves are the probable cause of Cramer's underperformance. As the stocks revert to the market's trend in the weeks after the show, Cramer's followers get hurt [See chart below]. Like any active-investing strategy, Cramer's advice must always be measured against the market return that his viewers could get in an index fund.
IT IS RARE THAT ANYONE BEATS the market over time, so there is no disgrace in the underperformance of Mad Money's stocks. The stocks featured in Barron's bullish stories did even worse than the market last year. ("Oops! We Missed the Mark in '08," Jan. 19)
Yet the last time Barron's inquired about Cramer's stock-picking, CNBC responded with cherry-picked success stories; lawyers; calls to Dow Jones executives; and an end to Barron's regular presence on CNBC. Cramer shouted to his viewers that we were know-nothings and assured them that his Mad Money picks had "killed" the Standard & Poor's 500 index. This time around, CNBC wouldn't let us near their headliner and said our questions were aimed at helping CNBC's less-watched rival, Fox Business News (owned by News Corp. , as is Barron's).
"You wrote a premeditated hatchet job to curry favor with your new bosses at News Corp.," said CNBC's Steel on Friday. "[Cramer] doesn't consider you a journalist."
The pre-show moves made by Mad Money stocks relative to the market were first observed by doctoral students at Northwestern's Kellogg School in a 2006 working paper. After hearing from an indignant CNBC, co-author Joseph Engelberg stopped labeling the moves "information leakage." When Barron's asked in 2007 about the pre-show moves we had found in Mad Money stocks, CNBC scrambled $100,000 worth of lawyers and sternly explained the broadcast lockdown procedures at the Mad Money set.
In the recent seven-month period, the pre-show runs are still the most dramatic thing about Cramer's stocks. We found that his bullish picks had risen 4% against the S&P in the two weeks ahead of his recommendation, while his bearish selections had dropped more than 7%. This action looks all the more interesting when compared with the pre-show activity in stocks that Cramer considers only when asked by a caller during the show's "Lightning Round." As the chart below shows, there are almost no market-excess moves before he tells a Lightning Round caller to Buy, while the Lightning Round Sells make but a fraction of the pre-show moves of previously prepared Sells.
MEASURING SUCH MOVES was easy, thanks to the tools available at EventVestor.com, a startup created by Wharton Business School and Merrill Lynch alumnus Anju Marempudi, with the advice of finance professors. Hedge funds and investor-relations firms are using EventVestor to study the returns of stocks around events like dividend cuts and earnings preannouncements.
So we got a record of the Mad Money recommendations from a source that Cramer endorses as the definitive way to track his performance. It is a trailing six-month database updated daily at TheStreet.com, the Website that Cramer brought public in the dot-com boom (see it yourself at MadMoney.TheStreet.com).
We then poured Cramer's data into EventVestor. Event-study tools like EventVestor aren't hard to understand. They simply track the performance of stocks over identical periods; for example, 10 trading days before through 45 trading days after each Mad Money show (as illustrated in the chart). You can leave out the impromptu advice he gives callers during the Lightning Round -- which Cramer has said shouldn't count toward his performance, even though the next-day stock moves show that Lightning Round watchers take him at his word when he tells them to Buy.
Looking at just the 650-odd recommendations Cramer prepared for the show's Discussion or Feature blocks between June and December, his bullish picks underperformed the S&P by about 3.5 percentage points over the 45 trading days after each show. His bear calls turned a slim profit of one point versus the market -- with all of the profit coming the day after broadcast, so viewers would do well to ignore Cramer's occasional urging that they wait five days before following his calls. You can even isolate the stocks of companies whose executives Cramer interviews and usually endorses -- those endorsed stocks dropped six points versus the S&P in the 45 days following the interviews. Considered separately, Cramer's Lightning Round Sell recommendations did better than those he prepared, while his Lightning Buys did even worse than those he prepped. [For charts of these results, and others, see Barrons.com.] It is reasonable to measure Cramer's stocks over such a relatively brief interval because -- as CNBC points out -- he isn't running a fund in which he reviews each position daily.
But the network and Cramer have alternatively argued that his picks are meant as long-term investments, so we also measured their performance from each show date through the end of the year. On that basis, Cramer's Buys finished five percentage points behind the Nasdaq and 10 points behind the Dow, while his Sells were one point less profitable than the Nasdaq but five points more profitable than the Dow.
Cramer bashers and acolytes typically argue in anecdotes. His critics remind you that he scolded a caller "No! No! No! Bear Stearns is fine! Do not take your money out!" just days before the firm collapsed in March. But boosters brag of his Oct. 6 market call on the Today Show, when he said: "Whatever money you may need for the next five years, please take it out of the stock market. Right now!"
That Oct. 6 advice saved investors "millions," said CNBC's Steel, by allowing folks to escape the market's 15% plunge through December. In fact, says Steel, that single piece of advice means Cramer beat the market, if you credit the 15% to his performance through Oct. 6. Of course, Cramer went on to make 800 more recommendations through December -- most of them Buys. Cramer would have saved investors even more, said Steel, had they put 20% of their assets in cash on Sept. 19, as he suggested. "Jim made two of the greatest prepared bearish calls of all time," crowed Steel.
We gamely worked through the details of CNBC's argument: Ending the measurements on Oct. 6 makes Cramer look worse, with his recommendations losing eight percentage points against the S&P. If you then spot him the Today Show 15%, as Steel insists, Cramer would finish the year seven points ahead of the market.
If readers don't buy CNBC's complicated argument, it has others. "Jim's advice is nuanced, complex and often qualified on either a future price or a specific market event," said Steel, who says that even Cramer's official Mad Money database misses nuances. It is kind of bizarre to hear the network impugn the Website that carries Cramer's endorsement as the record of "exactly what I say, when I said it, and how I feel about each stock now." He urges -- "passionately" -- that his show's performance be measured with those data.
When Barron's asked CNBC for their own preferred database of Mad Money recommendations, we heard something equally strange: The investment news channel keeps no track record of its stockpicker's Buys and Sells. "The show as it is currently produced," said Steel, "isn't set up to track every stock Jim mentions every day as if it was a fund."
Instead, Steel demanded that Barron's join him in watching six months of recorded shows so that he could decide whether Cramer really meant that viewers should buy or sell a stock. He said Cramer's Website had misinterpreted recommendations on four dates- for example, putting down a Buy recommendation when Cramer meant it sarcastically.
The Bottom Line
By most measures, Jim Cramer did worse than the market, but CNBC and the TV journalist have taken few steps to clarify his exact performance for his show's growing audience.In other words, CNBC wanted to debate its horse bets after knowing how the races ended. There is no way such a post hoc selection could be as credible as the record made at the time of each show (and before the recommendation's outcome is known) by Cramer's official Website. That would also be the time for Cramer to correct confusion in the record he tells viewers to rely on. Still, we recalculated Mad Money's returns without the four dates that Steel says had errors: Cramer's performance was precisely as bad without them.
The finding that Mad Money lags the market has been replicated using other records of Cramer's picks, too. University of Dayton finance professor Carl Chen used the third-party Website called MadMoneyRecap.com to study options-market trading in stocks that Cramer recommended. Chen found signs that the smart money bets against Cramer's Buy recommendations by using short-term in-the-money puts. Those bets could earn over 25% in a month, Chen concludes, at the expense of Cramer's fans.
CNBC's evasiveness about Mad Money's performance can't be attributed to Cramer, since the network wouldn't let us talk to the star. We were scolded that we didn't understand the mind of a genius. "Barron's and News Corp.'s repeated attempts to take Jim down have been a complete and utter failure," said spokesman Steel.
E-mail comments to firstname.lastname@example.org