| The company... reportedly made $2.1bn (£1.4bn) in profits from its pension fund in 2000 and 1999 - despite the fact that the fund has been losing money thanks to sliding stock markets.
The journalists would have us believe that accounting is a simple matter of counting the beans and printing the result, and if the number is wrong then somebody is a crook. Not so. The purpose of the sort of financial statements reported to the public and the SEC is to give investors a reasonable snapshot of the corporation's financial health at a certain point in time. A secondary, but highly desireable goal, is that these snapshots should be prepared in a comparable way from one period to the next, so can people can reasonably use them to assess a company's performance over time. OK, here is a stock portfolio owned by the GE Pension Fund. You're the accountant, and your job is to figure out what number we write down as the value of this portfolio. One legitimate number to use is the amount paid for it. That has the virtue of being a historically known fact. For some assets, and stocks look like a good candidate for this treatment, it is common to use "the lower of cost or market." Under that rule, we compute the value of the portfolio as if it were being sold today. We could use the closing prices on the last day of our fiscal year, and pretend we sold it all on that day, and add up the amount we would have received had we sold it. This number has the disadvantage that it is a lie, because the stock was not in fact sold. But it is a number that most people would say was a reasonable attempt to value the portfolio. Under "the lower of cost or market," we choose whichever of the two numbers we have now is lower. Actually, any of three numbers (cost, market, or lower of the two) is defensible as "the value of this portfolio." But notice that already there is no one number that is the "correct" number for the value of this portfolio. We can't just "count the beans." We have to make assumptions, and pick a number that makes sense for the purpose at hand. Actually, none of the numbers above make sense for the purpose at hand. We are attempting to value a pension fund so that we can tell people how this company is doing. The fact is, the pension fund did not sell the portfolio on the last day of the fiscal year, and it has no intention of selling most of the portfolio for years to come. What we would really like to have, so we can do this job right, is the closing price of each kind of share we own, on the final day of the fiscal year, for the next 20 or so fiscal years, when we're actually going to have to sell some stock. Of course, no one knows that. So to value this portfolio, the very best practice is to follow the example of Commander Spock, who said, "Mr. Scott cannot give me exact figures, Admiral. Therefore... I shall make a guess." And that is exactly what they do. It's "precision guesswork," in the sense that they will have their rocket scientists make predictions of the DJIA for the next 20 years, and munge that against how their individual stocks are doing relative to the DJIA, but when it's all rolled up, it amounts to a guess. It's an educated guess, but it's a guess. And that guess is probably a more accurate representation of the value of the portfoilio than any of the numbers (like cost or market) that are known today. In other words, we have just introduced a multi-billion dollar guess into the financial statements of the General Electric Corporation, and that is the best that humans can do. There is no crookery there... it is an honest attempt to tell the public what effect having that portfolio around has on the financial well-being of General Electric. But when it comes right down to it... it's a guess. If GE hires five accounting firms to do this, they'll get five different guesses. Hopefully they would all be in the same ballpark, but maybe not. Watch what happens if instead, these BBC journalists have their way. They say the company reported gains on the portfolio "even though stock prices have been sliding." What they've done here is adopt a market-price valuation for the portfolio, which sounds reasonable but in fact is a lousy way to account for pension funds in financial statements. If companies really did that, their reported earnings would bounce around like a yo-yo in response to stock market movements, even though no stock was in fact being sold. One could not tell whether the management of GE was doing a good job or a bad job, because the company's "bottom line" would be subject to wild fluctuations from one quarter to the next, based on nothing that had to do with the company. So these journalists are implying that something funny is going on at GE, when nothing like that may be happening. Certainly there is nothing wrong with GE not reporting the value of its pension fund as if it had all been sold on some date in the past (when it wasn't), and there is nothing wrong with GE trying to keep its reported earnings based on how the company is doing, not on the day-to-day fluctuations in the prices of shares held by its pension fund. The comment from Warren Buffet is more worrisome. Warren Buffet understands all this stuff, and he knows very well that the reported value of the pension fund is a guess. Warren owns lots of companies, and the reported values of their pension funds are guesses as well. What's he's saying is that -- in his judgement -- GE's is a pretty rosy guess. It is unlikely that GE executives took the output of the rocket scientists and arbitrarily jacked it up. Shareholder suits are too numerous and too expensive to mess with doing that. More likely they did hire competing teams of rocket scientists to value the portfoilio, and they picked the guess they needed to make the numbers for the year. Their intentions may not have been pure, but they were picking from guesses, so who is to say they didn't pick the right guess? It may bother people that there are guesses in accounting statements, but this is real life. No one knows for sure what that pension fund will be worth when the time comes to sell it. And it does no good to jack the company's bottom line around based on what it's worth right this minute, because that distorts the company's earnings. There just is no good, simple answer to this. It's a complex problem, with fuzz in all the reasonable answers. What we could probaby use less of right now is journalists playing accountant, using what they've learned from playing with QuickBooks to comment on the accounting practices of multinational corporations with subsidiaries in twenty different industries, a pension fund, and a financial services arm. Doing financial accounting in these kinds of enterprises is a lot like what you've heard about making laws and sausages: you wouldn't want to watch it being done. |
This is a perfect example of how accounting can become very complex, how simpler accounting can be very misleading, and how a lot of perfectly legitimate 'fudge-factor' gets into corporate books.