Selected writings by David Fiderer
When I was a corporate loan officer in the 1990s, some of my clients financed inventory with standby letters of credit, for one very obvious reason. An LC priced at 1% is cheaper than a loan priced at Libor plus 1%.
This banal insight — unfunded credit costs less than funded credit — explains why so much about the $187 billion “bailout” of the government sponsored enterprises makes no sense. The GSEs’ conservator, the Federal Housing Finance Agency, drew down $187 billion of taxpayer cash, which did almost nothing to enhance the companies’ operations or financial soundness.
The FHFA seems to have overlooked two critical distinctions. It overlooked the difference between cash losses and loan loss provisions under GAAP. And it overlooked the difference between a liquidity crisis and insolvency.
The cash draws were supposed to replenish the GSEs’ reported losses, which proved to be a chimera. We can look back with the benefit of hindsight at the 2008-2011 financials of Fannie Mae and Freddie Mac and see that accountants overestimated credit losses by a factor of 2X.
Because the FHFA drew down vast amounts of taxpayer funds to “pay” for accounting provisions that would later be reversed, the GSEs remained saddled with in senior preferred stock that paid a dividend rate of 10%, until that rate was revised upward.
Until mid-2012, the FHFA drew down cash to boost the GSEs’ book value from a significantly negative number up to something approximating zero. In the real world, if financial institution’s book equity is close to zero, it’s gone way beyond the tipping point. Under those circumstances, investors don’t really care about net worth; they care about the government’s commitment to support the company, which included purchases of their mortgage-backed securities.
The GSEs paid fees on the U.S. Treasury’s commitment to purchase up to $200 billion (later $400 billion) of senior preferred stock, which could be issued if the GSEs were ever unable to meet their current obligations. Until that date, there was no reason to draw down government cash for stock that paid out an annual 10% coupon.
By draining cash from the companies, the senior preferred dividends impeded the companies’ ability to build up retained earnings. The coup de grace was the FHFA’s decision to draw down $46 billion in taxpayer funds, so that the GSEs could pay senior dividends to Treasury. At the time, during 2010 and 2011, both GSEs had generated insufficient “earnings” to support those dividend payouts. Where I come from, a taxpayer “bailout” that (a) does nothing to fund company operations and (b) leaves the government with exactly as much cash as before, is oxymoronic.
We can walk through Fannie’s numbers, which show that it was in a situation akin to that of the Federal Housing Administration, which benefited from government guarantees. After 2007, FHA quadrupled its business volume to finance a mortgage portfolio that preformed far worse than those of the GSEs. Yet FHA never drew down a single dollar in bailout funds during the crisis years, and all signs show that FHA will continue to be self-supporting for the indefinite future.
The GSEs Were Not Like Wall Street
Contrary to popular myth, Fannie and Freddie did not face the same kind of liquidity crises that beset Wall Street banks in September 2008. The GSEs did not depend on overnight deposits, which can be withdrawn without notice. They did not require the massive liquidity needed to support trading of securities and derivatives conducted on behalf of clients. The GSEs did not own or insure exotic investments, like collateralized debt obligations, which were subject to margin calls.
Fannie and Freddie required less liquidity because their primary business line, guarantees of mortgage securities, was a form of unfunded credit. Those guarantees did not convert into a major cash drain in September 2008 because the residential mortgages did move in lockstep and all suddenly become 90 days delinquent. As of June 30, 2008, Fannie and Freddie’s serious delinquency rates were 1.36% and 0.93%, respectively.
The GSEs, like banks, financed long-term assets with short-term liabilities. But this duration mismatch was manageable. They funded mortgage loans, in part, by issuing 90-day notes with staggered maturities. But if the GSEs were unable to roll over those notes, they had the means to pay them down. Fannie, for instance, owned $418 billion in loans and $240 billion in short-term debt as of June 30, 2008. If necessary, Fannie could pay down half of its short-term debt by liquidating its cash equivalents and other assets guaranteed by the U.S. government. And it held about $260 billion in GSE mortgage securities, which could be pledged at the discount window the New York Fed.
Solvency Was Anybody’s Guess
Disaster was not imminent, but in September 2008 no one knew what knew future held. Defaults were accelerating everywhere and the size of the GSEs’ net worth was anybody’s guess. Using conservative loan loss assumptions, accountants inferred that the companies world never realize economic value for their deferred tax assets. Consequently, Fannie and Freddie recognized huge accounting losses, which were based on guesstimates about the future outcome of an unprecedented housing crash. As any banker will tell you, a bank’s solvency is based on a series of guesstimates — loan loss provisions, estimated prepayment rates — that are continually revised in both directions.
The bigger issue was the GSEs’ wherewithal to keep lending into a declining market after all the private players had walked away. So, to preempt any misgivings about doing business with two companies with negative equity, the federal government announced its bailout package.
From that point onward, there was no reason to draw down government funds to support the GSEs. Their government backing was akin to the government guarantee afforded the Federal Housing Administration, which quadrupled its business volume after 2007, while financing a mortgage portfolio that preformed far worse than those of the GSEs. Yet FHA never drew down a single dollar in bailout funds during the crisis years. And for the foreseeable future, FHA looks on track to be self-supporting.
Most sophisticated investors recognize that a bank’s solvency can change dramatically on a future date, when loans are taken off the books, and accounting provisions are reconciled with actual loan recoveries. And until that future date, there’s no reason to buy senior preferred shares prematurely, and pay a 10% annual dividend prior to that date of reckoning
Fannie Mae by the Numbers
Let’s walk through Fannie Mae’s numbers, which are substantially similar to Freddie’s numbers. At yearend 2008, Fannie’s single-family mortgage portfolio totaled $2.73 trillion. That $2.73 trillion was comprised of loans segmented by vintage in its Dec. 31, 2008 Credit Supplement. Five years and nine months later, in its Credit Supplement for Sept. 30, 2014, Fannie discloses the cumulative default rates by vintage. So, looking at the pre-2009 years, we can calculate a cumulative default rate of 6.6%, or about $181 billion of the $2.73 trillion total. Multiplied by a conservative loss severity rate of 37%, those defaults translate into cumulative credit losses of about $74 billion.
From 2008 and 2011, Fannie recognized single-family credit expense of about of $153 billion. That’s $153 billion (loss provisions plus foreclosure costs) versus $74 billion in actual credit losses to date.
Because Fannie’s credit losses were presumed to be so high during 2008 through 2012, virtually all of its deferred tax assets were presumed by accountants to be worthless. (Deferred tax assets measure GAAP loan loss provisions recorded before troubled loans are liquidated, when a tax deduction may be recognized.) By yearend 2011, $64 billion in deferred tax assets were offset by a deferred tax “valuation allowance” which represented a $64 billion reduction of shareholder equity.
If you appreciate the difference between cash items and non-cash charges, the FHFA’s decision to draw down taxpayer funds to supplant the loss of deferred tax assets seems especially foolish. Compared to all the other assets on Fannie’s balance sheet — anything from office supplies to real estate acquired through foreclosure proceedings — deferred tax assets are unique. Every other asset is convertible into cash. So, any reduction in value of those other assets translates into a reduced ability to pay down debt. But deferred tax assets are never convertible into cash, unless the company starts making so much money that it starts paying cash income taxes. Which is why, in the history of modern accounting, no deferred tax asset had ever been used to pay down a debt.
Beginning in 2012, the housing market had reached an inflection point, and prices started rising again. That year, Fannie started reversing those loan loss reserves and started earning profits. For fiscal year 2012, Fannie earned $17.2 billion in income before taxes, which was more than enough to pay down $11.7 billion in cash dividends, or 10% of the $117.1 billion in senior preferred stock owned by Treasury.
But 2013 proved to be an earnings bonanza, thanks in large part to prior period accounting adjustments. In 2012, Fannie’s single-family credit expense was about $1 billion. In 2013, some loan loss provisions from prior years were reversed, so that credit “expenses” added back $11.2 billion to corporate income.
Fannie earned $38.6 billion before taxes in 2013. But there was a double accounting whammy. Since it was obvious to everyone that Fannie might start paying income taxes in the future, a $45.4 billion “tax benefit,” which was the reversal of a valuation allowance, caused Fannie to report $84 billion in net income for that year.
Let’s do some math. Take $84 billion in net income 2013; then add back $30 billion in senior preferred cash dividends paid during 2009-2012. That would have been $114 billion available to boost retained earnings, which is pretty close to $117 billion drawn down by FHFA to “bail out” Fannie.
And even today it looks like Fannie’s loan loss provisions are excessively high. Remember, Fannie incurred $153 billion in credit expense during 2008-2011 versus $74 billion in actual credit losses to date.
So guess how much of that $84 billion in 2013 net income was used to boost Fannie’s net worth. Almost none of it. In mid-2012, the FHFA and the Department of Treasury decided that it was best for everyone if the GSEs paid Treasury cash dividends equal to 100% of net income. Which is why, in 2013, Fannie paid Treasury cash dividends totaling $82.5 billion.
That’s right. Most of Fannie’s earnings in 2013 were comprised of reversals of non-cash provisions from prior periods. But those non-cash gains were used to justify cash dividend distributions to the government.
None of it seems to make any sense. Under federal statute, the FHFA, as conservator, must, “take such action as may be necessary to put the regulated entity in a sound and solvent condition.” No one has ever given a cogent explanation as to how the original cash drawdowns, and the subsequent cash distributions, including a 100% dividend sweep, serves that goal.