What you missed in your previous post was the fact that Treasury has to approve any release from conservatorship. Why would they consent to give up the warrants? And why would they consent to release the companies before they meet some pre-set minimum capital level?
You also mistake the purpose of the warrants. They were never meant to only be collateral. In fact, that explanation is kind of absurd: if FnF were not in a position to pay back the bailout money then their shares would be mostly worthless anyway. The warrants instead were issued to drive the price of the stock down to make the companies appear to be in more trouble than they actually were.
I agree with your premise than once the seniors are cancelled, all that remains is how to recap the companies. But I can't just assume that the warrants will not be exercised. By cancelling the seniors and allowing a release from conservatorship, Treasury is giving up $193B in liquidation preference and all the NWS dividends. Why would they give up the warrants on top of all that? Treasury has shown by nearly all its past actions that it will take every penny it can.
I think the bolded part is likely to be moot. Watt's capital rule envisions far higher capital standards than those that were in place before conservatorship. Re-imposing those standards is terrible in terms of political optics: they weren't enough to prevent the companies from going under in 2008, why would they be enough for the next downturn?
Except for Treasury, you mean. Trust me: they don't care at all about how much money they have already gotten, they just care about how much they can get in the future.
When it comes down to it, removing your assumption of no warrant exercise breaks your entire model. Now shares only rise to $6 instead of $30, so to get that $48B in common equity you have to float 8B shares instead of 1.6B. This basically divides your final share prices by 5: $25 for FNMA, $15 for FMCC.
Now that doesn't sound so bad, but it still assumes that pre-conservatorship capital standards will be used, which is a very flawed assumption given Watt's proposed capital rule with higher numbers. Those updated numbers will require $100-125B to be raised to meet minimum capital levels. Now you need another $20-45B from that common equity to meet the same timeline, which is another 3.3-7.5B shares. Suddenly FNMA is in a range of $13-17, and FMCC $8-10.
You also are using $20B for annual income, but restoring junior pref dividends and adding more from your $20B in new junior prefs imposes a roughly $3B per year cost in dividends. Now instead of splitting $200 into $125 and $75, it's $170 split into $106.25 and $63.75. That knocks the numbers above down another 15%.
With rosy enough assumptions you can make almost any numbers work. The challenge is to come up with something that everyone that has power can agree to. Your plan involves Treasury taking it in the rear and that just isn't going to happen.