Down Arrow Button Icon

Who’s right about GE?

Accounting sleuth Harry Markopolos believes GE is engaged in accounting fraud so vast that the storied conglomerate will soon be forced into bankruptcy. Markopolos was the lone person to sound the alarm on Bernie Madoff’s Ponzi operation, and gained fame and credibility by turning out to be spectacularly right, raising fears he could be right about GE as well.

General Electric, which boasts a new CEO and a board packed with accounting experts, insists that Markopolos is both wrong and in cahoots with short-sellers. “This is market manipulation, pure and simple,” CEO Larry Culp stated in a press release. “Mr. Markopolos’s report contains false statements of fact, and these claims could have been corrected if he had checked with GE before publishing the report.”

As Fortune wrote in August, it has become the most fascinating standoff in Corporate America. At stake: the credibility and future of one of the icons of American business, as well as the ability of a lone-wolf accountant to wreak havoc on a huge corporation.

The debate started when Markopolos released his scorched-earth, 169-page report “General Electric, A Bigger Fraud than Enron,” on August 15, in which he asserts that GE is hiding $29 billion in long-term care insurance liabilities. By Markopolos’s diagnosis, GE doesn’t have long to live. He says he has submitted evidence alleging “accounting fraud far more serious than either the Enron or WorldCom accounting frauds” to both the SEC and U.S. Department of Justice under their “Whistleblower” programs.

Markopolos will only get the whistleblower reward if his sensational allegations are well-founded, and GE is indeed another Enron. But he’s also benefiting from a play orchestrated to make himself big money even if he’s wrong—and that gambit calls his motives into question. Markopolos has disclosed, in the report and in interviews, that he gave the results of his still-secret investigation exclusively to a hedge fund that planned to short GE stock in advance of publication. In exchange, the hedge fund shared its profits with Markopolos. The fruits of their partnership could be big: On August 15, the day of the report’s release, GE’s stock dropped 11% to a low of $8.01, erasing $8.9 billion in market value. It has since recovered somewhat, closing Wednesday $8.51. The fund’s identity is still a mystery, as is the size of Markopolos’s windfall.

There’s no arguing that GE has been anything but a long-running disaster. Since the legendary CEO Jack Welch retired in 2001, its stock has dropped by 78%, erasing $250 billion in market value. Welch’s successor, Jeff Immelt, orchestrated ill-fated acquisitions in power (Alstom) and energy (Baker Hughes), and presided over the meltdown of GE Capital, which choked on risky real estate acquired during the turn-of-the-century bubble, and a gigantic, loss-plagued long-term care insurance (LTC) portfolio that’s still sapping its finances. Immelt’s successor didn’t last long, and investors are now pinning their hopes on Culp’s sterling record in building a powerhouse at Danaher, an industrial conglomerate that thrived just as GE, the colossus that created the template for success, was failing.

Markopolos’ charges are especially shocking because he’s attacking not the old regime that created the mess, but a new leadership team and board packed with people boasting both strong reputations and long expertise in insurance. Besides Culp, the lean, 10-member board contains five current or former CEOs, including two insurance veterans: James Tisch, chief of Loews Corp., owner of CNA, a provider of long-term care coverage; and Paula Rosput Reynolds, retired head of Safeco Insurance. Also onboard are former CFOs Tom Horton, who filled that post at AT&T and American Airlines, and Catherine Lesjak, who held the top finance job at Hewlett-Packard.

Heading the board’s audit committee is Leslie Seidman, formerly America’s chief accountant as chairman of the Financial Accounting Standards Board (FASB), the body that sets the bookkeeping rules for America’s publicly-traded corporations. One person familiar with Seidman’s role at GE says that her reviews of the numbers, and especially those from long-term care insurance, are so rigorous “it’s as if she’s eating nails for breakfast.”

After carefully weighing Markopolos’s bombshell charges versus the numbers that GE, supervised by state regulators, has produced, this reporter concludes that the picture presented by GE is more accurate. (Markopolos didn’t agree to be interviewed for this story, and GE decline to make executives available for comment.) To understand how we reached this conclusion, it helps to understand four things: How GE got into this mess, how insurance industry accounting works, what Fortune believes Markopolos gets wrong, and why ultimately Culp may be able to generate enough power to get GE out of this mess.

How GE got into this mess

It’s important to understand that the GE Jack Welch created wasn’t just a maker of jet engines, turbines and other industrial equipment. What Welch and later Immelt considered its jewel was GE Capital, the financing and investment arm that made venturesome bets in real estate, commercial lending, and insurance.

Long-term care insurance sales boomed in the 1980s and early 1990s. Young and middle-aged Americans in large numbers bought policies to cover nursing home and home care expenses, to kick in when and if sickness or frailty rendered them unable to live independently decades in the future. GE mostly acted as re-insurer for carriers such as Allianz and MassMutual that had sold the policies to customers. GE took the risk off their hands by booking the premiums, paying the claims, and forecasting the trends for both, as well as investment income, that determine annual profits or losses. In the early years, the business appeared highly lucrative. When its members were young, GE took in loads of premiums that far exceeded its claims, and forecast that the big revenue stream, coupled with strong returns on the pool of capital accumulating from the premiums it retained, would exceed claims far into the future. Those actuarial predictions enabled GE to book big profits on its long-term care portfolio in the 1990s and early 2000s.

But GE’s forecasts proved way off the mark, as did the projections for most of the industry’s big players. It reinvested far too little of those surpluses, and counted far too much as profits. The size of future claims depend on four principal factors. The three non-financial variables are “mortality,” or the customer’s longevity; “morbidity,” the estimates of the number of members who will become sick enough to require nursing home or home care, and the severity of their illnesses; and “lapse rates,” the share of members who stop paying premiums and drop out––the more terminations, the lower the future cost of claims. The fourth factor is the “discount rate,” the anticipated annual percentage return on the investment portfolio backing the reserves.

As it turned out, what went wrong with GE’s assessment of where these variables were heading recalls that old multiple-choice exam answer: “All of the Above.” Members are living several years beyond the lifespans predicted when the early policies were sold. The rise in the incidence of Alzheimer’s disease has forced millions more Americans into nursing homes than previous morbidity estimates predicted. And terminations, which GE had predicted at 3% or more annually, are running at 1% or less. As if it couldn’t get worse, the two-decade decline in interest rates culminating in today’s near-record lows crimped income from investments.

GE couldn’t escape the whirlpool. Its policies, like most in the LTC business, are “guaranteed renewable,” meaning they can’t be cancelled. To raise premiums, insurers must gain approval from state regulators, and authorities frequently reject the requests even when carriers or reinsurers are losing money and the increases are justified by the economics. (In an added complexity, it’s the company that sold and services the policy, not the reinsurer, that’s tasked with requesting the increases.)

Many insurers were guilty of allowing the shortfalls in their insurance reserves to grow to dangerous proportions. But GE ranked among the worst in failing to recognize the that it needed to make gigantic additions to its reserves. In early 2018, then-CEO John Flannery shocked investors by announcing that GE needed to fill a gaping, $15 billion hole in GE’s insurance portfolio, declaring his “deep disappointment with the magnitude of the charge.” Flannery lost his job six months later. Says Albert Meyer, an forensic accountant who heads Plano, Texas, investment firm Bastiat Capital, “It’s remarkable that the adjustments to reserves weren’t measured and consistent over the years to avoid these huge charges and contributions. Unless, of course, management prevailed to report better earnings and cash flow in the hopes of better days ahead, days that did not come.”

For more than a decade, GE’s long-term care franchise has been in “runoff” mode, meaning it’s doing no new business, and simply collecting premiums, and paying claims, for legacy customers who generally bought policies long ago. Today, GE still insures 342,000 customers at an average age of 77; it paid out around $600 million in claims in 2018. It’s now in the late stages, when the cash outflow from annual claims far exceed incoming premiums. The question is whether, even after pledging to add $15 billion to its portfolio, GE has enough money in the kitty to bridge the gap between those premiums and claims in the years to come. GE and its regulator say yes. Markopolos claims that GE needs three times that $15 billion bailout, and that coming clean, and funding the full deficit, means bankruptcy.

To determine which side is most convincing, it’s important to understand that GE, like all insurers, is required to follow the two accounting regimes imposing extremely different requirements.

Making sense of an accounting clash

The statutory accounting regime required by the states, or “STAT,” follows nationwide rules set by the National Association of Insurance Commissioners (NAIC), as well as requirements of the jurisdiction where the insurer in domiciled—in GE’s case, Kansas. Its numbers aren’t the ones you see on a company’s balance sheet and income statement. Those are governed by the second system, Generally Accepted Accounting Principles, or GAAP, whose rules are set by FASB. Put simply, STAT establishes more conservative safeguards than GAAP.

The STAT details for how the insurance products are performing aren’t easily available. The STAT financial reports are filed in the insurance commissioner’s office in the state in which they’re domiciled. Wall Street analysts seldom consult them. While the thick paper filings in yellow covers for property and casualty insurers have long been called “yellow perils,” notes insurance accounting expert Neal Stern, chief of consultancy Greenmeadow Associates, the bulging statutory filings for life insurance carriers, a category that includes LTC, bound in blue, might be called the “blue perils.”

Both STAT and GAAP mandate that companies hold adequate reserves, but their methods of calculating those reserves are different—with STAT requiring greater caution. Reserves aren’t actual big pools of money. Instead, they’re liabilities that represent the difference between the estimated value of the claims the company faces in the future, and the the premiums and earnings they expect to collect while awaiting the claims. In the case of STAT, the amount of those reserves must, over time, be offset dollar for dollar by an investment portfolio on the asset side of the balance sheet; that’s the pool of money backing the policies. Today, it’s about $25.4 billion for GE in long-term care. The STAT rule establishes a “discount rate” that estimates the future rate of growth of the reserves. The assets in the portfolio must be big enough to generate the cash needed to make up the difference between the annual claims and premiums.

Each year, GE is required to perform what’s called an “asset adequacy test” under STAT. GE, like all insurers, uses all four of the main variables, morbidity, mortality, lapse rates and the future estimated return on the assets, as well as projections on future premiums. But STAT applies strict methodologies to the way GE uses those factors to estimate the future cash outflow from claims and inflow from premiums that determine whether it needs to add to reserves and bolster its portfolio.

STAT sets limits on estimates of “lapse rates,” the proportion of customers who will stop paying and no longer require coverage, and caps assumptions for improvements in health for the elderly that might reduce the numbers entering nursing homes. Many states restrict the the extent to which insurers can assume that they’ll receive approval for future rate increases. STAT does not permit insurers to set their own discount rate. Instead, it establishes the number using its own formula based the current market rates. For example, GE’s STAT discount rate is now in the 4.5% range, well below the GAAP number.

An especially rigorous test is called the “New York 7,” named for the state that launched it, but applied nationwide. The idea is to determine if an insurer can reinvest cash from maturing bonds in the future at sufficient returns to cover claims in the years ahead—a potential problem in periods of falling interest rates. Under its methodology, the carriers, GE included, run seven different scenarios assuming various patterns for future interest rates: “Test three,” for example, assumes rates rise one percentage point a year for five years, then retreat by one point a year for the following five years; others assume falling rates in the early years.

If the insurer satisfies six of the seven tests, but doesn’t generate sufficient cash under just one, it needs to boost its reserves, and hence follow on by lifting investments so that the returns on the larger portfolio now cover the worst-case shortfall. That’s a template for the caution that rules in STAT-land. Say the asset deficiency test establishes a deficit, from changes in cancellation projections, estimates of more Alzheimer’s sufferers entering nursing homes, or satisfying a New York 7 test. That means the gap between the value of future claims and premiums has risen. STAT is all about ensuring that the portfolio is plenty big enough to cover future claims. If it isn’t, as exposed by the asset adequacy test, the insurer’s obligated to boost both reserves and investments so that the extra cash generated each year by the bigger pool of investments makes up the difference.

But the insurer doesn’t have to provide all the cash right away, and as we’ll see, GE is contributing the funds from its own shortfall in annual installments. “When an insurer has a deficiency, the insurance regulator may allow it to make it up by adding cash to its portfolio over several years,” says Greenmeadow’s Stern. A deficiency under STAT does not cause a loss on a carrier’s GAAP financial statements. But it’s still a problem. That requirement has forced GE to devote cash to a no-profits, runoff business, rather than using that money to rebuild its core industrial franchise.

Under GAAP accounting, the insurer also performs a review to ensure that future cash flows will match expected claims. It goes by a different name, “loss recognition testing.” But unlike STAT, GAAP does not impose methodologies on how GE measures the four prime factors that determine what those future benefits and revenues will be. That’s left to GE’s discretion. In the past, GE, like many other carriers, regularly under-estimated the gulf between its premiums and investment income, as well as the rising trajectory of payments to nursing homes, assisted living facilities, and home care providers.

Under GAAP, insurers are instructed to use their “best estimates” for mortality, discount rates and the other variables, based on both past trends, and how their actuaries view the future. But companies are obliged to adjust their numbers by something called a “provision for adverse deviation,” or PAD. The PAD effectively means that GE has to add a cushion to its forecasts based on the chance that, say, interest rates will be lower than their best estimates.

When an insurer falls short in the loss recognition test, it’s obligated to increase the reserves shown on the balance sheet. That results in a charge to earnings. By contrast, failing a STAT test does not cause a hit to GAAP profits. Instead, it mandates that GE put more cash into its investment portfolio. So it’s GAAP that governs reported earnings to investors, and it’s STAT that determines the size of the investment portfolio (and reported earnings to the regulators).

What Markopolos doesn’t account for

Because STAT is more conservative than GAAP, the STAT reserves required by the state regulators are generally a lot higher than the GAAP reserves on the official balance sheet for reporting to shareholders. That’s the case now with GE. And that difference is critical in evaluating the Markopolos case. Markopolos claims that under a new rule, GAAP reserves need to climb to the level of STAT, causing a big loss and endangering the entire company.

In January of 2018, GE disclosed that its GAAP loss recognition test showed a $9.5 billion deficit in reserves. As a result, it raised its GAAP reserve liability by that amount, and took a $9.5 billion hit to pre-tax earnings––a tremor that hastened Flannery’s exit. The increase almost doubled the GAAP liability from $10.5 billion to $20 billion. At the same time, GE disclosed that its STAT reserves, once again, calculated on a more conservative basis, were underwater by much more, $14.5 billion. That change in STAT reserves had no effect on its GAAP earnings, but required that GE boost its LTC investment portfolio by $14.9 billion, from $15.5 billion to $30.4 billion, to cover its previous under-estimates of future claims. The Kansas Insurance Department gave GE seven years to contribute the cash, and so far, it’s put in a total of $5.4 billion, with $9 billion to go.

Those big shifts left GE with $20 billion in GAAP reserves, and $30.4 billion under STAT. Markopolos claims that the new FASB rule issued in August of 2018, scheduled to go live in either Q1 2021 or Q1 2022, will require that GAAP reserves equal STAT. According to Markopolos, hardships imposed by that measure, Accounting Standards Update No. 2018-12, spell disaster for GE. As he states on page 19 of his report, “By 2021, GE must equalize GAAP and SAP [STAT] for existing reserves.” (Markopolos doesn’t mention that that rule could take effect a full year later, the most probable timetable according to industry experts.) As a result, he states in the introduction to his report, GE will suffer “a non-cash GAAP charge which accounting rules require.” He asserts that this $10.5 billion “charge to earnings will “result in a devastating, $10.5 billion hit to GE’s already thin shareholders’ equity.”

Markopolos is incorrect. “Nothing in that FASB rule requires that STAT and GAAP be the same,” says Stern. Stern’s view is echoed by all the accounting experts I’ve interviewed, none of whom have professional ties to GE. The new FASB rule on calculating GAAP reserves imposes requirements that have nothing to do with raising GAAP to the level of STAT, although it may coincidentally bring them closer together, and will not cause any hit to GAAP earnings. Today, GE can set its own discount rate based on returns for its current portfolio, and estimates of the rates at which it can reinvest the principal from maturing bonds in the future. The FASB rule would remove that discretion, and require that carriers all adopt the same rate based on a uniform benchmark, the current yield on investment grade, single-A rated corporate bonds.

At first hearing, that sounds dangerous. The single-A bond with the same average maturity as the GE portfolio, around 9 years, yields roughly 3.0%. But GE’s current discount rate, by Fortune’s estimates, is in the 5.7% range. GE has disclosed that every one point difference between its current discount rate and the new benchmark will necessitate a $4 billion increase in reserves. So if the shift happened today, and it’s most likely two years away, GE would need to boost its reserves by around $10.8 billion.

“But the FASB rule won’t cause any charge to earnings for insurers,” says Stern. The reason: In the past, if GE or another carriers added to reserves because interest rates dropped, lowering the projected returns on its portfolio, that increase in reserves caused an equivalent pre-tax charge to earnings. Under the new rule, the reserve increase will become a debit to a different account that’s part of shareholders’ equity called Accumulated Other Comprehensive Income (AOCI). It’s an entry that contains capital gains on securities still on the books. (This accounting phenomenon, given its first three initials, should not be confused with the Congressional phenomenon from the Bronx.)

As interest rates have dropped, the value of GE’s bond portfolio has risen sharply. As of June 30, GE harbored gains of $4.8 billion in marked-to-market, available-for-sale securities lodged in AOCI, and because yields have fallen since then, I estimate the current number at around $5.6 billion. For GE, the new rule offers a provision that softens the penalty. GE will be able to offset the charge from the rise in reserves by the subtracting the gains banked in AOIC. Hence, if the change came today, GE would book a charge to equity of approximately $5.2 billion (the $10.8 billion from increasing reserves, less the $5.6 billion in AOIC), or $4.1 billion after-tax. That’s around half the hit that Markopolos predicts. While that still would be a painful blow, it’s not nearly enough to force GE, with its $116 billion in annual revenue from industrial businesses, to seek bankruptcy protection.

FASB designed the change to set a uniform standard across the industry instead of letting insurers continue to set their own discount rates. The charge to equity is non-cash, and won’t much bother investors because it doesn’t signal any change in the cash inflows and outflows GE is already predicting. In other words, it doesn’t point to a change in the underlying economics of GE’s, or any other insurer’s, long-term care franchise. “Going to a 3%, single-A rate ‘discount rate’ doesn’t change the bonds an insurer holds in its portfolio, which might be yielding an average of 5% or 6% because they were purchased years ago, and mature gradually,” says Stern. “It’s not as if FASB came along and said, all your bonds are now yielding 3%. And it doesn’t imply any change in projected cash receipts or claims in the future.”

It’s the second part of the new FASB rule that poses a risk that’s challenging to predict. Today, under GAAP, insurers can “aggregate” all of the profits from the years they issued policies that are showing surpluses, meaning that estimated future cash flows exceed claims, and subtract the combined losses for all the years showing deficits. If the total profits exceed losses, reserves are adequate; in the opposite case, the carrier is obligated to raise its reserves to bring the two totals into balance.

When it goes live, most likely in 2022, the FASB rule will require that projected cash flows for each individual year show a profit. For each annual block of business that’s underwater, the insurer must raise reserves sufficiently to show a projected surplus. Today, if over the past seven years a carrier shows profits in 3 years, and losses in 4, but a surplus overall, its reserves are adequate. But after the change, it will need to increase reserves to breakeven on all of the three years showing losses, and won’t get any immediate offsetting credit from the four years of projected profits.

According to the insurance experts I interviewed, there’s virtually no public information indicating profits and losses estimated for individual years. That makes the impact on GE, and everyone else, hard to predict for outsiders. It’s one of the ironies of the GE-Markopolos melodrama that the sleuth makes no reference to this potential reserve-inflator.

The aggregation issue raises a new long-term care mystery for investors weighing the prospects for GE and other insurers.

Quantifying the long-term toll on GE

GE is already suffering a heavy burden from the insurance saga. The aforementioned STAT asset adequacy test required GE to shift $14.5 billion from its other businesses to support insurance, an area it’s exiting that will never make money. GE has already added $4.5 billion ($3.5 billion in 2018) in cash to the investments backing LTC, and will contribute the remaining $9 billion through 2024, at the rate of $1.8 billion a year. “That money has to come from cash from operations,” says Meyer of Bastiat Capital. Last year, GE’s cash from operations amounted to $8.4 billion, meaning that it would have been $11.9 billion or 42% higher if it hadn’t added $3.5 billion to reserves.

Even if GE is right and it doesn’t face a huge boost in reserves, the $1.8 billion a year going into a blighted business will hurt its future performance. “If you’re an investor, you want to think about the prospects for the core business,” says Stern, “you want the money to go into places that make money such as jet engines or medical equipment, not a place where you get no return.”

On the industrial side, GE lags in labor costs, working capital, overhead and other metrics that are obvious to Culp, and he’ll aim to raise GE’s performance in all of those areas to the excellence of star performers such as United Technologies, Honeywell and his alma mater Danaher. “I have to believe they can turn around the industrial side,” says accounting expert Jack Ciesielski. “It’s all about the basics of managing people, plants and cash better, and improving marketing and market share.”

In this reporter’s view, Culp’s goal should be getting out of long-term care, soon and for good, by handing the franchise to a rival insurer. It’s clearly an option he’s considering. “We’ve got a number of strategic options over time,” he told investors in March. Considering the industry’s history of heavy losses, finding a taker will be a tough task. GE might have to add billions of dollars in extra reserves as a sweetener. But dumping LTC would be the ultimate riposte to Markopolos, who’s claiming that GE’s book is so toxic no one would take it, and that the end game is bankruptcy.

Nothing is certain in the shifting sands of long-term care. Reaching a judgment on GE’s future depends on your view of management’s integrity. GE now has a new team of seasoned actuaries and insurance specialists running long-term care. Director Seidman, the former FASB head, has scoured the books. Culp spent four months on the board studying the financials before becoming CEO. That lowers the odds that the new leadership has missed a time bomb that will destroy GE.

The best hope is that GE dumps this deadweight, even if the cost is big (though hopefully not Markopolos-big), clearing the way for Culp to work his magic. The second best hope is that Culp is right and LTC holds no more terrors.

In that case, he could build GE back into an engine with enough power that the deadweight becomes an easy lift.