August 15, 2019 was a rough day for General Electric’s (GE) stock. Harry Markopolos, the guy who unraveled the Enron accounting fraud, released a 169-page report claiming that GE was a bigger fraud than Enron. The gist of the report was that GE was hiding a bunch of long-term care insurance liabilities and that if they reported the liabilities correctly, GE would take a charge to earnings that would “result in a devastating, $10.5 billion hit to GE’s already thin shareholders’ equity.”
GE’s stock closed down ~11 percent for the day, erasing ~$8B from GE’s market cap.
I hesitate to wade into arcane accounting debates since I studied philosophy instead of accounting in college, but it should be obvious that, for some reason, I can’t avoid discussions about long-term estimated liabilities.
The crux of Markopolos’s contention is that GE must bring its GAAP reserves up to match their STAT reserves. STAT reserves are established based on guidance put out by the National Association of Insurance Commissioners, supplemented by the laws of whichever state jurisdiction an insurer is domiciled in, while GAAP is, well, GAAP.
To massively oversimplify this, STAT reserves are more conservative as they mandate the discount rate instead of letting the company set the discount rate. As a reminder, the discount rate is the estimate of the future growth rate of the assets set aside to meet future claims. There are quite a number of other factors that are used to calculate STAT vs GAAP reserves, but the big difference is the discount rate.
GE’s STAT discount rate is now in the 4.5 percent range, which is well below the GAAP numbers they’re currently using.
Markopolos’s GE Claim
This leads us to Markopolos’s claim that under new GAAP guidance, GAAP reserves must match STAT reserves by 2021 or 2022. Currently, GE has $20B in GAAP reserves vs $30B in STAT reserves. Markopolos claimed in his report that the new FASB rule is scheduled to go into effect in 2021 or 2022 and that GE will have to take a non-cash charge to earnings to equalize them.
Where this gets weird is that most accounting experts agree that the FASB rule that Markopolos cites does not require raising GAAP reserves to the level of STAT reserves and it will not cause any hit to GAAP earnings.
The reason this is true is that the offsetting entry for the reserve increase will hit a shareholder equity account called Accumulated Other Comprehensive Income. It totally bypasses the calculation of net earnings from the income statement.
The table above is from GE’s most recent 10-K. It’s the third page of their income statement. You can clearly see that Other Comprehensive Income (Loss) is calculated after net earnings.
Markopolos’s Bigger Error
The bigger error is Markopolos’s premise that a change in reserves and its impact on shareholder equity is what will sink GE. Monkeying around with reserves on a balance sheet doesn’t matter in the short term. You can have negative book equity on a balance sheet but still have a giant market cap and a very bright future.
The simplest example of this is a startup. Imagine there’s a startup that’s been in business for 5 years. They haven’t turned a profit yet and through their annual losses they’ve depleted their initial Paid in Capital (for you accounting nerds out there, the Paid in Capital entry will remain the same but there is an offsetting retained earnings entry that is negative). Their balance sheet will show a big fat zero for shareholder equity. Now imagine this startup lands a massive contract. Let’s say Amazon wants to license their technology for $10 million per year for the next 20 years. The owners don’t want to wait 20 years to collect the money so they talk a bank into giving them a loan for $50 million. The company receives the loan then immediately pays out a $50 million dividend to the owners.
The owners then throw a massive party in celebration of their good fortune.
The $50 million is still a liability, but the offsetting asset (the cash) is gone, so their book shareholder equity is now negative $50 million. Not just zero dollars, negative $50 million. Is this startup now bankrupt? No, of course not. They have $10 million per year coming in from their technology license, which is more than enough to pay the debt service on the $50 million loan with quite a bit left over.
GE is certainly the opposite of a startup, but what matters is how much cash their core businesses can throw off going forward. If GE can throw off enough cash to service their remaining insurance liabilities while continuing to invest and expand their core industrial businesses then they will continue to have a viable business.
Markopolos does have a point though…
While it’s weird to think he missed the boat on the accounting rules as badly as he did, almost as if he had some ulterior motive for releasing the report, he does have a point about understating reserves. While it won’t matter from a net earnings standpoint, it does matter from a cash flow standpoint. If the reserves are understated, and I’m pretty sure that they are based on their return assumptions even though they are perfectly fine per GAAP and STAT standards, at some point GE will have to make good on those obligations.
If it costs them $2B per year from operating cash flow to do so then every investor will have to adjust their fair value estimate downward to account for the ongoing drag. While GE would be best served by offloading the residual liability to an actual insurer, it’s unlikely at this point that they’d swap a certain liability (taking on debt to fund the transfer of the liability) with a squishy liability they can kick the can on.