6 reasons why Wall Street analysts mis-estimate company earnings

companies continue to beat Wall Street estimates

It’s earnings season again.  As companies report their results for the quarter, investors watch carefully to see whether they match, surpass, or fall short of, the consensus estimates of Wall Street securities analysts.  As has been the case since the bull market began over two years ago, a majority of companies are beating the consensus figures–some by a mile .

Why don’t analysts do a better job of forecasting?   

I have six reasons (five of them today, the final one tomorrow):

1.  Some companies are very hard to forecast, because they’re so complex.  They may have a lot of divisions, many suppliers and a large variety of finished products that they sell.   

Alcoa, the first of the big publicly traded US corporations to report each quarter, is a case in point.  You’d think that the main variable for Alcoa would be the price of aluminum, which you can see every day in commodities trading.  …but no.

Alcoa can get its raw materials from many sources.   Some sources are joint ventures that Alcoa’s a part of, which often have complicated profit-sharing arrangements that are not publicly disclosed.

Miners deliberately change the grade of ore they mine, depending on price, to maximize the life of the ore body.  Ore can be processed in different locations, some company-owned, some not, depending on the price of electric power and other factors.  So expenses are hard to gauge.

On top of all that, differing accounting conventions at each step of the way can play a big role in what cost figures eventually appear on the income statement.

In a case like this, it’s hard to imagine anyone outside the company having a good handle on what reported earnings will be.

2.  Some companies demand that analysts stay close to “official” guidance.  CEOs understand that having earnings per share that exceed the consensus is a good thing for their company’s stock  (a positive earnings surprise), and that missing the consensus can be a very bad thing  (a negative surprise).  So most firms that issue next-quarter guidance to analysts give out numbers they believe they have an excellent chance of beating.

Some firms go further than that.  They make it clear they will punish any analyst who deviates from guidance more than a little bit.

What can a company do?

Lots of bad stuff.  From descending order from worst to not-as-bad, the company can:

–stop using the analyst’s firm for any investment banking services, like underwriting debt or equity offerings (which is likely to get the analyst fired),

–not appear at industry conferences the analyst’s firm may organize, or send the investor relations guy instead of the CEO or CFO,

–avoid using the analyst to organize the company’s visits to powerful investment management companies,

–refuse to give the analyst access to top company officers to ask questions about operations,

–delete the analyst from the queue to ask questions during earnings conference calls (petty, it’s true, but not unusual).

There have been cases of stubborn analysts who have stuck to their guns in print, against company wishes.  Their stories  usually end badly.  For most, they’ll show the number the company wants in their written reports and “whisper” their best guesses to clients.

3.  Some analysts aren’t good with numbers.  That’s not always a fatal flaw.  The analyst may have other pluses–a deep knowledge of the inner workings of an industry, an ability to  sense new trends quickly, or very good contacts with the companies, or customers and suppliers.  Or maybe they just know where to take clients to lunch.  After all, when you get down to it, sell-side analysts are paid for their ability to generate commission business  for their trading desks and to attract/keep investment banking clients–not necessarily for having the best numbers.

4.  Sometimes analysts don’t do the spreadsheets themselves.  An assistant does them instead.  Analysts may spend half their time on the road visiting companies and clients.  Much of the rest of the time, they’re doing the same thing on the phone.  And the analyst may not be so great with numbers, anyway.  So that task is relegated to an assistant.

An aside about assistants:  in my experience, many analysts pick assistants who are articulate, look good in a suit and are smart–but not too smart.  Why?  A sell-side analyst may earn 10x-20x what an assistant does, in an industry that suffers a severe downturn in profits about twice a decade.  The analyst may worry that having a top-notch assistant means the analyst becomes a cost-cutting victim during recession.

5.  Many experienced analysts have been laid off over the past few years.  Doing good estimates doesn’t require a genius, but it does require some training and experience.  A lot of that has been lost on the sell sidesince the Great Recession began.  In addition, as I suggested above, the assistants who have been promoted into the principal analyst jobs may not all be the brightest crayons.

That’s it for today.  I’ll write about #6, making really dumb mistakes, tomorrow.