A few posters seem to be asking, "What if the Glazers do not write off their portion of the PIK notes?"
Well, if we take GCHQ's assumption that the notes would have to be paid off evenly, that is, any payment towards them would have to be spread across all the notes regardless of the identity of the holder, then you'll find that the net effect would be exactly the same as if the Glazers did write off their portion of the PIK's.
In explanation of this, look above at the 'Scenario B' spreadsheet. This shows how soon the PIK's could possibly be eradicated if the Glazers wrote off their portion. Look at the 'Total PIK' level for year 2011; after the £95m payment is made and the interest added for 2010, the total of outstanding PIK debt in 2011 is £91m.
Now, consider what would happen in the same situation but instead with the Glazers' portion of the PIK notes still in place.
The Total PIK debt year 2010 would be £200m, and once interest is added at 16.5% this would increase to £233m, divided 20% - £46.6, and 80% - £186.4 between the Glazers and third party holders.
The £95m payment would then similarly have to be divided, 20% - £19m, and 80% - £76m, which would reduce the total debt to a split of 20% - £27.6m, and 80% - £110.4m between the Glazers and third parties.
Now, the Glazers have just paid themselves £19m for their 20% holding of the PIK notes, so this extra cash can be paid back in.
Split it 20% - £3.8m, and 80% - £15.2m and the new PIK totals are; 20% - £23.8m, and 80% - £95.2m.
Again, the Glazers just paid themselves £3.8m, which can again be paid back on the PIK's. You get the gist;
Split 20% - £0.8m, and 80% - £3m the new totals would be 20% - £23m, and 80% - £92.2m.
This process can continue indefinitely until you're calculating just pence, but if you stop there for the sake of sanity, look at the total PIK debt for the remainig 80%; £92.2m; remarkably close to Scenario B's £91m in the same year; i can assure you that the £1.2m difference comes only from rounding the figures up or down through various stages of each calculation; the net effect to the PIK's would be exactly the same regardless of whether or not the Glazers write off their portion.
So by buying the 20% in 2008 the Glazers have effectively already written off that portion of the debt, because an IOU to oneself is absolutely worthless; either way, whether they write off the PIK's now or continue to hold them for tax purposes or whatever, the net effect to the remaining 80% is equal.
Edit.
This is purely just a mathematical concept, perhaps anders, GCHQ or someone else could comment and expand upon its practical application?
Yes, it is an iterative process that (as described above) asymptotically tends to the cancellation scenario outlined in your spreadsheet.
I think Andersred has already commented on the iterative process.
There is a pretty simple solution which, from a practicable point of view, kills the pointlessly recursive process of using ever diminishing pik returns to reduce the total outstanding pik balance still further.
The Glazers simply need to inject cash (x*dividend) into RFJV immediately prior to the pik redemption dates such that their share of the total amount redeemed ((1+x)*dividend) is equal to the amount they injected.
Solution: x =25%. (from x*dividend=.2*(1+x)*dividend)
So, for an annual dividend of, say, 30m, the glazers add 7.5m, the total pik remeption becomes 37.5m, and a week later auld Malc receives 7.5m in the post that matches the amount stolen from his piggy-bank a week earlier.
Now, if you allow for this in your spreadsheet, it will not exactly mirror the cancellation scenario but pik elimination occurs at the same time, the amount paid by the club in dividends will also match up. The total interest bill will be a bit higher though.
There are other solutions; the Glazers could use their pik proceeds as a balancing fund to adjust the amount they want to take in dividends- it is, afterall, discretionary subject to a maximum of .5*CNI.
As for the rest of your post; there are other scenarios. In fact, a whole multitude of them.
The Glazers might simply pocket the redemption proceeds thus rendering reduntant our ruminations on the vexations of recursive algorithms.
Why?
For one thing, though the club is an asset appreciator, it hasn't been a good provider of the other component of equity return- cash income in the form of personal dividends- and this is mainly down to the various covenants.
Holding the piks entitles them to a discrete dividend stream of 20% (reportedly) of an annual dividend amount that they can control (subject to a limit). With a dividend of 95m (the carveouts a.k.a. the 'Ronaldo money' plus a bit) the Glazers can reduce their pik indebtedness by 95m and collect 19m at the same time. They can no doubt find use for this money elsewhere and probably can offset the amount against losses elsewhere. They might even use the 19m to pay back the 10m loan from the club. Who knows? THe point, of course, is that the 20% of the pik proceeds is theirs to spend without restriction unlike the annual dividends and carveouts.
For another, the actual cost to them of not using their pik redemption proceeds to further unwind the pik is not as big as you might expect.
In your cancellation scenario, the Glazers cancel their share of the pik, bringing the outstanding balance down from 200m to 160m. The club saves about 110m in pik payments. However, the Glazers, in cancellation, miss out in 60+m of uunrestricted pik dividends. There is also a cost (with cancellation) in terms of missed interest expense relief of about 20m or so.
So the difference (for them at least) isn't too big over the entire period that the pik remains on the book. You might also want to factor in the Glazer's need for cash in arriving at the likeliest scenario.
Another option is for the Glazers to cancel their share of the piks after the carveout of, say, 95m is taken. The 19m (16m of the 'Ronaldo money') windfall will come in handy I'm sure.