The reporting on the main elements of the Freddie and Fannie rescue plan is converging as the content of official briefings leaks out.

The stunner, which contradicts preliminary reports, is that the preferred shareholders in the GSEs will take losses. The Wall Street Journal reports that dividends on common will be eliminated and those on preferred will be suspended (Bloomberg, Reuters, and the New York Times were less specific, but indicated that preferred shareholders would suffer).

Justin Fox summarized why this outcome, of wiping our or otherwise damaging the preferred shareholders, had been considered highly undesirable:

Lots of small and mid-sized banks in the U.S. have, with encouragement from regulators, built up big holdings in Fannie and Freddie preferred stock, which they use to satisfy their capital requirements. If Fannie and Freddie preferred shares become worthless, a lot of banks will become insolvent. Which, with the FDIC insurance fund already being depleted by bank failure, could end up costing taxpayers a ton.

Now admittedly, the preferred shares have already inflicted sizable losses to those who bought them when issued. For instance, Freddie Mac's Z preferred was sold for $25 and has traded in the $10 to $14 range recently. But even at that level, its rich dividend (at these level, the yield is 15%+) is a support for the price. Adjustable rate preferred is trading at a bigger discount, 80+%. Where do they go once that prop is removed? Readers who can provide insight are encouraged to speak up.

In the meantime, let's do some first-cut back of the envelope calculations. The face value of GSE preferred was $36 billion. While the market values have taken a beating with the bailout talk, most banks would have end of June prices reflected in their latest financial reports. Since I don't have access to historical prices or the mix of adjustable versus fixed rate, let's assume market value across all issues was $18 billion as of end of June. Even with the dividend suspension plan, the preferred will still have some option value. This is a complete stab in the dark, but say the shares fall to 1/4 their current price. So the incremental damage is $13.5 billion. Now that doesn't sound all that bad across a whole lot of banks (boy, have we gotten inured to big numbers). But one can of course argue the contrary case, that an equity hit of that magnitude reduces bank lending capacity by roughly $180 billion.

Now to do this analysis correctly, not only do you need better numbers on 1) what losses have the banks taken already on Freddie and Fannie preferred and 2) how much more of a hit is likely, but you also need to know 3) which banks have a significant exposure relative to their capital. Even if the aggregate losses are not awful in a macro sense, they could have the effect of tipping a seemingly disproportionate number of banks over the edge.

Nevertheless, inflicting damage on the preferred sharholders was seen as so unlikely that John Dizard of the Financial Times devoted his last two articles to the merit of investing in GSE preferred shares. He regarded a cram-down as a non-starter (this is one reason we steer clear of discussing investment ideas here. You can do good analysis and still have your head handed to you. But in fairness, if you had taken his advice as a trade and gotten in and out fast, you would have made a very nice turn).

From the Financial Times:

In the now overcrowded world of investing in distressed securities, the standard strategy is to pick the "pivot" issue...that is the problematic part of the capital structure. Securities with seniority above that of the pivot get paid out, securities below that get wiped out or have their value seriously impaired. The idea is to buy the pivot at a good price.

In the past couple of weeks, it seemed that the entire US economy had a pivot security or set of securities: the preferred stock issued by the government-sponsored entities, or GSEs. Fannie and Freddie, the Sodom and Gomorrah of "public/private partnerships", sold about $36bn of non-cumulative preferreds to the banks and the public, with the aggressive support and encouragement of the Treasury and the GSEs regulator.

Last week I wrote about these preferreds; my position was that if or, rather, when the Treasury had to recapitalise the GSEs with new, senior preferred issues, it would be a really good idea from the taxpayers' point of view to leave the old preferreds in place while wiping out the value of the outstanding common stock.

I thought that saving Fannie and Freddie's preferreds would support the entire asset class of preferred stock. The banking system needs to raise several hundred billion dollars of equity, and preferred stock is the lowest-cost way to do that in the public markets. While some sophisticated investors could distinguish between preferreds issued by a sound bank holding company, and preferreds issued by the overleveraged F&F, international investors and domestic retail investors would not have the data or analytics to draw the distinction.

The alternative, as I see it, to recapping the US banking system with preferreds is some form of direct government investing in the equity of banks or bank holding companies. That would be even more expensive to the taxpayers - as in at least 10 times more expensive.......

I called up Andrew Senchak, the vice-chairman and co-director of investment banking at Keefe Bruyette & Woods, which specialises in bank securities...

As he says: "It is true that there is no direct link between the GSE preferred issues and that of the banks, but they are in the same galaxy. Given that, there is almost no incentive not to keep the GSE preferreds in good shape. If there is a recapitalisation of the GSEs [by the Treasury], you can achieve the public policy end [of limiting 'moral hazard' by wiping out the value of the common]. I am not sure how much new bank equity has to be issued . . . it could be anywhere from $200bn to $400bn."...

And yes, I agree that it is likely, if not certain, that if the GSE books were marked to market, the asset value would not be there to support the preferred issues. There is, though, a real value to clapping to keep Tinkerbell alive here: you get a banking system that can finance a recovery.

As a reality check I called Jim Grant, of Grant's Interest Rate Observer, and the author of the forthcoming Mr Market Miscalculates . He comments: "The alternative to preserving the value of the GSE preferreds? Prayer? Remember that a lot of that paper is held by the same banks the authorities would love not to fail."

So, with the market's affirmation, and the agreement of one of the Street's flintiest sceptics, I'm sticking to my position: the GSE preferreds should survive.

Narrowly, Dizard is 100% correct. The GSE preferreds will indeed be preserved. But if the news stories pan out, that will be cold comfort to their owners.