THE BLOG
10/10/2012 11:50 am ET Updated Dec 10, 2012

Jack Welch Knows A Thing Or Two About Dubious Numbers: GE Capital Edition

A little-remembered episode in General Electric history illustrates just how dangerous it is for former CEO Jack Welch to accuse the government of manipulating unemployment numbers.

Flagged for The Huffington Post by an anonymous tipster, the episode suggests that five years of sterling GE earnings were themselves the result of aggressive data-massaging. It suggests that nearly all of those earnings were a mirage -- as was Welch's reputation as a manager.

Welch on Wednesday doubled, tripled, or quadrupled -- I've lost count -- down on his claim that the drop in the unemployment rate in September, to 7.8 percent from 8.1 percent, was the result of the Bureau of Labor Statistics manipulating the data for political gain.

Many observers who only vaguely know Welch as a respected business legend expressed shock that he would stray so far from the reservation of rational thought. But a few close students of his oeuvre, including Barry Ritholtz, noted that Welch and the company he used to run, General Electric, have long had a reputation for earnings manipulation, for making the numbers meet or beat expectations quarter after quarter.

Many have pointed to a 2009 settlement between the Securities and Exchange Commission and GE, in which the company agreed to pay $50 million to settle charges it had fudged its numbers for several years with dodgy accounting techniques including fake sales of locomotives.

The trouble with this example is that it mostly covers the years just after Welch left the company, when CEO Jeff Immelt was struggling to maintain the company's sterling reputation for almost never missing earnings targets.

There is another, little-remembered example of dubious data, though, that comes straight from the Welch era. It was discussed in detail in a 2005 Barron's article by Jonathan Laing, called "Jack's Magic: Revising A Legacy."

In the article, Laing points out that GE's earnings in the last five years of Welch's tenure rose from 72 cents a share to $1.37, a 90 percent increase. That gain cemented Welch's reputation as a business legend and, more importantly, helped secure him a retirement package (which he later renounced) that included a Manhattan apartment, use of a corporate Boeing 737, country-club fees and much more.

The trouble with this near-doubling of GE profit was that it was due almost entirely to socking away too little money to protect GE Capital's reinsurance unit, Employers Reinsurance (ERC), to cover future insurance claims. After Welch left the company, Immelt was forced to bulk up ERC's reserves by $61 billion, or 61 cents a share -- effectively wiping out all but 4 cents per share of the earnings gains seen in the last five years of Welch's tenure.

In other words, Laing argues, if GE had not left ERC woefully under-reserved, GE's earnings would have grown by just 4 cents per share, or 5.6 percent, in the last five years of Welch's run. A far, far uglier performance, not the stuff of legends.

It is a little unfair, perhaps, to single out Welch and GE for massaging earnings, when it's fairly standard procedure in corporate America. A recent study by economists at Duke and Emory found that 20 percent of public companies "manage earnings to misrepresent economic performance."

But then, Welch is practically begging to be singled out.

In a 2006 letter to the editor responding to Laing's article, Welch defended himself, saying ERC reserves were simply responding to conditions in the broader economy. When times were good, there wasn't much need to add to reserves. When times were bad, starting in 2001, reserves suddenly needed padding.

"It is foolish -- in fact, it is one of the oldest tricks in the book -- to isolate one factor in a large, complex company and twist it to make a point about an entire enterprise," Welch wrote. "It is worse still to take such convoluted logic and use it to impugn someone's integrity (in this case, mine!) and that of an entire organization."

You may now take a moment to chew on the delicious irony of this statement from Welch, in light of the fact that he is now impugning the integrity of the government, accusing it of manipulating just one number -- the unemployment rate -- to a still-not-good 7.8 percent, while leaving U-6 unemployment, payroll growth and other numbers at ugly levels.

In his 2006 letter to Barron's, Welch also insisted that Laing had failed to get in touch with him to discuss the article. Laing responded, in his own letter to the editor:

I could argue with the many specious points made by Mr. Welch, but let me focus on just one: I indeed did try to reach him, through General Electric. I was assured by the company that he had received my contact information, was apprised of the thrust of our story, and might or might not reply. He didn't.