I wish you health, happiness, and prosperity in the New Year!

Thanks for the wonderful thoughts.
Do try to remember me when it's time fro the Feste award.
"Vote early and vote often" is my motto!

I plan to retire most of this margin by May 09 or cover it with sell
orders. I do not intend to sell more than margin exposure for 10 years.

I too am long Berkshire with as much margin as I'm comfortable with.
Actually, just a tiny bit more than that, but I sleep too much anyway.

Rather than timing your sales, I'd recommend pricing them.
I don't plan to sell any of my margined shares till they hit 130,000 per A,
unless I have other reasons to be more uncomfortable with the margin level.
If you're more conservative, wait for $120,000, which is the low-end
price that I would expect today based on historical discounts to IV.
Better yet, don't sell until the market price is greater than
the "conservative" setting on the intrinsivaluator site, which
admittedly might be a long wait.

Here's an interesting calculation for you:
Imagine a portfolio of nothing but Berkshire for funding a retirement,
purchased today at $92,000 per A share.
If the price is low, you don't want to sell---you're willing to use
a tiny bit of margin for your living expenses so that you can wait, and
your next sale as after a new high. You sell enough to pay off the
small margin amount. Presumably a year or so of living expenses is a
small fraction of your portfolio's value, so the margin level will be small.
If you can't wait for a new all-time high, you at least promise
yourself to wait till a new one-year high.
(actually Berkshire tends to hit long strings of new all-time highs, so
in the past it has statistically been worthwhile to wait till it drops
3% from its all time high before selling, but I digress).
Further, you're going to keep working till the stock hits $130,000.
No sales or withdrawals before then, in other words---just hold your breath.
In this scenario, I estimate you could withdraw 10% of the original
dollar amount every year for the next 30 years and still have more
dollars worth of stock than you started with. Well, with a 99.3%
probability in my simulations. How's that for a safe withdrawal rate?
At a nice round $10,000 per A share withdrawal rate (never increasing
with the portfolio size or inflation), or 10.87% of initial portfolio
value, you would have over a 98% probability of a higher dollar balance
in each of years 10, 20, and 30.
This is a very long and involved simulation that I've described before,
so I won't go into the boring details. Suffice to say that the
assumptions going into it are not crazy: I assume that Berkshire
gradually loses its ability to outperform the broad US market, and that
the range of discounts to IV is somewhat broader than historically seen.

Jim