Hollywood studio execs are relearning the painful truth that it pays to under promise and over deliver. Otherwise you might end up with Godzilla. Some companies take a similar approach to investor relations: underplaying earnings estimates in the hope of encouraging a pop in their stock when official earnings beat expectations. In their own way, however, they too have created a monster.
‘Whisper’ estimates are unofficial earnings forecasts that circulate among traders and investors. While not new, they have become a key stock price catalyst, especially for hot technology and other earnings – sensitive stocks – those particularly favored by momentum traders.
Try asking phenomenally successful computer maker Dell. Its stock got clobbered earlier this year after its record first quarter earnings matched analyst consensus estimates (found on First Call, I/B/E/S and others) but missed the whisper number by a whisker. Similar seemingly bizarre curiosities have occurred at a growing litany of companies including Hewlett-Packard, Intel and Cisco Systems.
In effect, for some stocks, the whisper has become the true market expectation. So-called ‘whisper forecasts’ are now reported in the financial press and commonly discussed on investor web sites. Most IROs will hear about them from buy or sell-side contacts. You can rule one out if your stock keeps running up after a better-than-consensus earnings release. If not, some sleuthing may be in order. The whisper may come as a warning.
To understand the nature of whisper estimates and where they originate, consider the term as a rather unquantifiable buzzword subject to interpretation. Some estimates are whispered anonymously and thrive in internet chat rooms and bulletin boards. Others are offered up by analysts to favored clients but not reflected in their published estimate. Then, interlinked with the first two, there is the kind of whisper engendered by companies with a pattern of beating consensus estimates.
While the internet has formalized the whispered earnings phenomenon on sites such as Silicon Investor and Briefing.com, internet whispers are about as chimeric as you can get. ‘No one knows where they come from,’ comments Stan Levine, director of quantitative research at Thomson Financial. Remarkably, however, internet whispers are generally closer to the mark than the published analyst’s consensus.
Levine provided data for a pioneering study done by university researchers that compared whispers culled mostly from web sites and electronic bulletin boards with traditional analyst forecasts from January 1995 to May 1997. Not only did the study conclude whispers tend to be more optimistic and accurate than consensus, it also suggests whispers have arisen, in part, because investors have come to believe that analysts underestimate earnings. ‘Whisper numbers generally act as a better proxy for the market’s expectation of earnings,’ says Mark Bagnoli, visiting associate professor of finance at University of Michigan and one of the paper’s authors.
The mass media likes to imply that Wall Street trades on inside information and has pointed the finger at sell-side analysts as the shadowy force behind whisper numbers.
Chuck Hill, director of research at Boston-based First Call, says that notion is a ‘crock’. ‘There is a number floating around, but it isn’t coming from analysts,’ says Hill, who had 20 years experience as a sell-side analyst with Kidder Peabody and Quantum Associates. According to Hill, most analysts really do call them as they see them; he asserts that their reputation and compensation are largely predicated on estimate accuracy.
Color of money
‘An analyst’s reputation is their most valuable commodity,’ says Hill. While he acknowledges that estimates can sometimes be colored by investment banking relationships within securities firms, Hill is firm on this point: ‘Whisper numbers are in no way created by the main thrust of the analyst community.’
Hill recalls that as an analyst, he put ‘sells’ on investment banking clients and ‘never heard boo’ from the investment banking department or the company. By the same token, however, a cooperative relationship undoubtedly exists between sell-side analysts and the corporations they follow. One way analysts may nurture that relationship is by rarely publishing sell recommendations. Another is to follow guidance and publish estimates within a range the analyst thinks the company would like to see.
Indeed, subjective analysis by clients plays a large role in analyst rankings, such as those found in Institutional Investor. Analysts may in fact earn a better recommendation by giving clients privileged information that is not broadcast.
David Powers, a sell-side analyst at St. Louis-based Edward Jones, suggests another reason for the poor showing of consensus estimates against whisper numbers. He says the emergence of whisper estimates has less to do with IRO-to-analyst communication than with analyst-to-client communication. ‘As you approach the release date, clients tend to call and ask about your confidence in those numbers,’ says Powers. ‘If you feel a certain way, it’s easier to communicate that verbally as opposed to actually crunching all the numbers and coming out with a new published number.’
Indeed, analysts don’t simply change their published numbers because the process often requires the approval of a committee at the brokerage. In the face of uncertainty, the instinct is to go low. Some analysts will simply low-ball their numbers for companies they are enthusiastic about so it can be said the companies beat estimates.
Clearly, there are problems with consensus estimates since most companies beat them. Regardless of where it originates, the bias in consensus numbers is becoming increasingly recognized. Rumors aside, so-called ‘whispers’ are merely a rational response being built into expectations.
Conservative pattern
While analysts spin their numbers, First Call’s Hill places the primary blame squarely on IR departments. ‘They have created their own monster,’ says Hill. ‘Beating the numbers is okay in the short term. But once you establish a pattern of beating them quarter after quarter, you build expectations and create a whisper number.’
That appears to be the trap Dell slipped into after having regularly beaten consensus expectations by some three cents per share in the past eight quarters.
‘Consistently guiding the Street to estimates below the actual numbers is dangerous IR practice,’ comments Richard Wines at Georgeson & Co. ‘Manipulation of the consensus estimate may work in the short term, but over the long haul you damage your credibility. Generate artificial surprises and they will come back to bite you when you cannot meet expectations.’
‘You always want to meet or beat estimates by a small margin,’ adds Evan Klein, managing director at Carson Group’s San Francisco office. ‘Sell-side analysts love those kinds of companies because they are the most predictable. Predictability and visibility get you a higher multiple.’
Still it is hard not to pity the IRO hunched over the phone talking down earnings estimates. If his or her company misses those targets, even by a penny, billions of dollars of market capitalization can evaporate within days as PE ratios shrivel. Admittedly, accountants usually give IROs a range of earnings estimates to work from. Software companies are especially prone to ‘hockey stick’ quarters with sales forces closing business at the last moment.
While advising IR departments to manage expectations closely, Lou Thompson, CEO of Niri, admits the temptation to be conservative is strong. ‘In this earnings-sensitive environment, it is better to understate and overperform than to do the opposite,’ says Thompson.
Certainly, pleasant things happen to your stock after a positive earnings surprise. In fact, outpacing analyst expectations has become a major ingredient in corporate success in the 1990s. The bull market has made the task easier. But the market has become wise to the tactic and it is losing effectiveness. For some companies, it has become a liability.
Taming the beast
Taking the steam out of whisper numbers means breaking the pattern. Those companies that have painted themselves in a corner will probably have to weather an adjustment process and eat a few earnings ‘disappointments’ where earnings meet consensus, but fall short of the pattern. If a company’s earnings closely match analyst consensus for several quarters, investors will note the consistent pattern. That gets factored into multiples. Meanwhile, the whisper number disappears.
If a company really wants to put a stop to a whisper number, it can issue a pre-announcement. These have become more common as the deterioration in global economies has dampened the pace of earnings growth. Sometimes, however, the cure can be as bad as the disease, with subsequent volatility the rule.
Otherwise, if a company’s earnings will not reflect estimates for any reason, the situation can be managed by understanding how individual institutions will react under different earnings conditions. Stock surveillance specialists can advise on the way valuation will be affected and identify the institutions that could buy the stock based on its new price.
Then there is the delicate question of how much earnings guidance to give. Say too much and an IRO could face a lawsuit. Say not enough, and a negative earnings surprise can spark chaotic volatility. In general, sell-side estimates should be kept within a narrow margin. The task is made more difficult by spotlight-hungry analysts who establish their turf as either bears or bulls, following the pack rather than company fundamentals.
Some companies give a lot of guidance. Others let analysts draw their own conclusions. Intel, for example, gives guidance at each quarter’s end for the following quarter on issues including earnings, gross margin and tax rates. From that point on, the investor relations team lets the numbers speak for themselves.
‘We do not comment on analyst estimates whether they are written or whispered,’ says Tom Waldrop, an Intel spokesperson. ‘If we feel the need to update that guidance, we will do so based on our own guidelines and not whisper numbers.’ Waldrop says Intel’s guidance is as accurate as possible. ‘The [whisper] phenomenon is not something that emanates from us,’ says Waldrop.
Arise, uncertainty
At the end of the day, if a company has a volatile earnings stream, uncertainty among analysts will arise and the word will spread to institutional sales forces. You can’t always stop the whispers, but companies aren’t helpless either. How and what managements communicate over time has a major effect on whispers.
Whispers typically appear around the time a company goes into its quiet period, making them especially hard to deal with. Once in the quiet period, IROs cannot comment on quarterly results, but they can remind investors that guidance during the pre-quiet period was stable.
‘Providing uniform and consistent guidance as you move through the quarter helps prevent the generation of a whisper number,’ says John Lawlor, IR director at Ottawa-based telecom equipment producer Newbridge Networks. ‘Whisper numbers arise only if there is uncertainty about the company actually delivering the published earnings estimates.’
Still, one can attempt to be consistent yet remain at fortune’s mercy. Over the last fiscal year, Newbridge had ‘disappointed’ three quarters in a row. Yet in Newbridge’s case, unanticipated business reasons were largely to blame, including an underperforming acquisition. Nevertheless, a fertile environment for a whisper number had arisen. Newbridge bit the bullet in February, when it warned results would fall short of analyst expectations. Shares plunged 28 percent in one day.
Lawlor was doubly challenged in Newbridge’s fourth quarter when he was suddenly confronted with yet another surprise. Three working days before the official earnings announcement, a top Wall Street firm launched coverage with an estimate a penny over consensus. Poof: a whisper number appeared. On June 3, despite meeting analyst expectations, Newbridge shares fell 12 percent.
There will always be surprises for investor relations officers. That’s partly why the job is fun – it keeps people on their toes. Momentum investing, for example, is becoming so common that it is losing its efficacy, and will give way to new styles. In the meantime, the Asian crisis has profoundly shaken the corporate profit outlook. The days seem numbered for across-the-board positive news on earnings. Those IROs who have carefully guided their flock of analysts up the earnings slope may soon have to bring them back down again.
