This commentary was first published on SmartMoney.com on February 6, 2002

In the wake of the Enron scandal, investors are all of a sudden focusing intensely on the integrity of corporate accounting. So self-righteous pundits and politicians are elbowing each other for space at the head of the lynch mob demanding new regulations to enforce more standardized disclosure.

I don't for a moment seek to excuse the kind of accounting fraud in which it appears Enron participated. Yet, at the same time, the howls of protest about garden-variety corporate accounting that's merely complex and ambiguous strike me as insincere. I'm reminded of the story of the old lady who complained about an obscene phone call — "It was horrible! For 45 minutes this awful man said the most terrible things!"

Well, if it's so horrible, just hang up the phone. And if a company's accounting is too arcane, just don't invest.

But the reality is nobody really wants to hang up. Both investors and corporations have always liked a little complexity and ambiguity in their accounting, especially when it earns a red-hot growth company a red-hot price/earnings multiple. Who cares if the numbers are a little convoluted so long as everybody profits? Perhaps it's what the new-age psychologists would call a "co-dependency" — a destructive relationship that nonetheless meets everybody's needs.

A perfect example of this kind of co-dependency is Tyco International. A Bermuda-based global conglomerate knitted together from hundreds of mergers and acquisitions both large and small, Tyco appears to be a growth machine, racking up rapidly ramping revenues and earnings year after year. But at the same time, with all those acquisitions, Tyco's financial statements may be among the most complex and opaque of any public company's...ever.

Last week investors were outraged when stories circulated that Tyco had failed to disclose details of over 700 acquisitions worth $8 billion made over the last three years. The company defended itself by claiming that there were just so darn many of them, each one was effectively immaterial! And why not? Is anyone really surprised? Hey, all those droolingly admiring articles about Tyco in the mainstream press over the past couple of years have all fawned over its hyper-efficient acquisition process capable of digesting a new company every week.

The panic induced by this and other issues last week didn't abate until Tyco Chief Executive Dennis Kozlowski announced that he and his finance chief would personally buy a million shares of Tyco common stock to demonstrate their confidence. Like J.P. Morgan quelling the panic of 1907 by buying when everyone else was selling, Kozlowski's grand gesture restored confidence, at least for a while. But this week Tyco has fallen to new lows, as investors have become concerned that — accounting issues or no accounting issues — Tyco is becoming so tarnished and beleaguered that it may not be able to meet its ongoing financing needs.

Tyco finds itself facing a potential death spiral — induced by issues that may have no substantial reality. Why? Because Enron has forced investors to face up to their co-dependency with Tyco on complex and ambiguous accounting. That's what fuels the conglomerate game. And it always has, all the way back to the "go-go" growth stocks of the 1960s. The stakes are bigger today, but this is nothing new.

Conglomerates thrive by creating a track record of rapid and sustained earnings growth. They don't have to go to increase their earnings the hard way, by creating new products or opening up new markets; they do it by buying other companies, and using accounting conventions to absorb them in a way that creates the appearance of growth.

Here's how it works. Conglomerate GrowthCo trades at 100, and has earnings of $5 per share (giving it a price/earnings ratio of 20). Now GrowthCo buys TargetCo, which trades at 100 and has earnings of $10 per share (for a P/E of 10). If the two companies are the same size before the acquisition (for the sake of a simple example), then after the acquisition GrowthCo's earnings will be $7.50 per share, the average of the two companies' per share earnings. All it took was a stroke of a pen, and earnings appeared to grow 50%, all thanks to the arithmetic of per-share accounting. Now imagine what you could do with so many acquisitions that you can't even disclose them all.

This effect is what analysts mean when they say that an acquisition is "accretive to earnings" — and according to Tyco's annual report, immediate accretiveness is a requirement for every acquisition it considers. But does accretion of earnings through acquisition mean that an acquiring company is growing? Not at all — it just means that its stock has a high enough P/E ratio so that the arithmetic works out like this.

With a high P/E — and a high stock price — an acquiring company has to trade away relatively little of itself to absorb the target company's earnings. But with a low P/E it would have to trade away so much of itself that the same acquisition could be dilutive — the opposite of accretive.

And how does a stock get a high P/E to begin with? By demonstrating growth potential. So let's see if you've got this straight: The acquirer manufactures per-share growth by having a high P/E, and it gets a high P/E by manufacturing per-share growth. Isn't accounting wonderful? See what I mean when I call it a co-dependency? And do you see why investors have gone for years without asking too many questions about Tyco's tangled books?

The legendary hedge-fund manager George Soros sees it more philosophically. He cites the co-dependency between a conglomerate and the investors who bestow upon it a high P/E ratio as an example of his "theory of reflexivity" — the way financial markets don't just reflect the world around them, but turn around and influence real events in the very world they reflect. But Soros points out that reflexive phenomena in markets can turn out to be both virtuous and vicious cycles. And that's precisely what conglomerates have learned periodically throughout the past 40 years, and what Tyco is learning right now.

As long as the market bestows a high P/E on a conglomerate, its cost of capital is so low it can keep on growing through accretive acquisitions and justify that high multiple. But if, as in a game of musical chairs when the music suddenly stops, something happens to reduce the conglomerate's P/E, then it will stop growing. And that prospect reduces its P/E even more — and so on in a vicious cycle that can become the kind of death spiral that Tyco faces today. In fact, Tyco's current plan to split itself up into several smaller companies is exactly the kind of unwinding process that you would expect to see when the cycle turns vicious.

I have no idea if Tyco's accounting issues will end up being deeper than this. For all I know, next week's headlines could have Tyco involved in something worse than even Enron. But I doubt it. Where there's smoke, there's not always a smoking gun. Tyco will probably turn out to be just another conglomerate that lived and died by the sword of acquisition accounting, and by investors' desires to believe in a growth story as long as other investors believed in it, too.

We'll all be sadder and wiser for a while. And then in a couple of years, it will start all over again. And no new regulations about accounting will be able to do anything about it.