First off, Bigshan pts out in Column 1 and 2 valuation, that Cash and
investments is not = in value to sh'holders as op earnings, seems to
equate it to just excess cash on a BS and earning no real return.
Any further commentary on this aspect of BRK?

A first comment is that I believe that the "two column" approach is a
great way to understand Berkshire, but at best a so-so way to value it.
The reason is that no valuation of Berkshire can really be "good enough"
unless it understands how an insurance company works.
In particular, the two-column values published in the annual letters
implicitly assume that there will never be any underwriting profit or loss,
for the simple reason that they entirely ignore the fact that it's an insurance firm.
That being said, when today's price is far below intrinsic value measured
any number of different ways there isn't much point quibbling about
the most accurate process. I have moved to simpler and simpler valuation
methods over the last couple of years because fancy stuff isn't needed—for now.

To your question, you're right. The ongoing existence of the firm requires
that there always be a big pile of cash earning only very modest returns.
It can't all be deployed, and can't be paid out as dividends, so it
will always be a drag in that the earnings on those particular assets
will always be lower than the earnings on others.
So, one of these dollars of cash in the hands of Berkshire by itself
doing nothing is not worth as much as a dollar in the hand of a shareholder.
There are a couple of mitigating factors: first, as mentioned above,
Berkshire is an insurance firm and (unusually) one that definitely
earns an underwriting profit average through the years.
Thus, the overvaluation of a simple two-column approach which values
the cash as highly as anything else is offset by the undervaluation
because there is no estimate of ongoing average underwriting profits.
It's the insurance subsidiary that means the cash has to be there,
so it's fair to include the upsides of that subsidiary.
A second mitigating factor is that cash does earn interest in normal years.
The current ZIRP (zero interest rate policy) won't last forever.

Along the same lines, the publicly listed shares held by Berkshire
are not worth quite as much as the same publicly listed shares if
they were spun off to Berkshire's shareholders. The reason is the
extra level of taxation on the dividends. This is a pretty small
effect, almost certainly made up for by the fact that the people who
work at Berkshire seem to be better than average at picking stocks.
The two-column approach has no place to include any profits from
buying low and selling high, which does happen from time to time.
e.g., the PetroChina investment.


Also, curious about what ppl on this board think about Tangible Book Value vs Book Value when valuing BRK.

A rather glib but vastly true viewpoint is that it just doesn't matter
because Berkshire has so many assets that don't show up in book value at all.
Another thing to note is that a surprisingly large fraction of the
goodwill on the books is from a single purchase of BNSF.
That purchase turned $X of cash into less-than-$X of tangible book.
But was value destroyed thereby? Isn't BNSF worth more as an
operating business than the scrap value of the rails and cars?
I think most people would agree that BNSF is worth at least what
was paid for it, so the goodwill from that is real value.
Besides, if it were still a listed company it would likely be on the
books at a very much higher value, probably tracking the other big rails.
So the acquisition created a goodwill asset you think you might have
to mark down when it would otherwise have created a huge unrealized
capital gain if it were not consolidated which we'd never question.
UNP's share price is up 128% (not 28%) since the day before the BNSF acquisition
was announced making the 30% premium to undisturbed price seem pretty minor!
Though this discussion addresses only BNSF, much of the other goodwill is similar:
the business in question is really worth more than its book value, not less,
since the goodwill generally came from the gap between tangible assets of the
purchase price of an operating company which is worth more than its original purchase price.
In short, no, I don't think a tangible book adjustment is necessary or even prudent.
If you track tangible book through time you'll see a big downwards
discontinuity because of the railway purchase which seems rather perverse.

Which moves us on to the issue of book value as a value metric at all.
As mentioned above, at huge discounts it doesn't matter what valuation
methodology you use, you'll still get the right answer.
But, price/book is not a very good metric. Short term rate of change
in book value is a not-bad but generally understated proxy for short
term change in value, but over time it's a problem. You can't compare
any two firms by price/book unless they are in substantially the same
business and have structures and balance sheets of very similar type.
You can't look at the 1.26x of Berkshire today and say it's cheaper
than the 5.35x of Coke. Similarly, and very much less obviously, you
can't say that the 1.26x of Berkshire today is 19% cheaper than the
1.55x of Berkshire five years ago because the structure of Berkshire
has been changing, again in large part because of the rails and utilities.
So, tracking price/book of Berkshire over time and saying it's cheaper
than its average is a very iffy proposition.
Price/book is a very crude yardstick, best avoided beyond quick
generalizations or when valuations are extreme.
I imagine Berkshire as it is currently constructed is worth
over 1.5x book, maybe 1.7x, but that may change as Berkshire changes.
The only good way to get that number is to value it a better way then
look at what multiple that value is of current book per share!

Finally, a lot of ppl seem to argue that a multiple of about 10x seems
reasonable for value of BRK's operating subsidiaries...

Indeed. Bear in mind that people are talking about 10x pretax earnings.
If you pencil in a rough tax estimate of 35%, that equates to a P/E of 10/(1-.35)=15.4x.
Personally I like to err to the conservative side and might use 9x.
That corresponds to a P/E of about 13.8x.
The average US listed firm since 1937 in the average year has traded at a P/E of
about 13.86x on-trend after tax earnings, so it certainly seems like a sane range.
In reality 10x pretax is probably justified for the simple reason that
Berkshire's operating companies are definitely of slightly higher
average quality than the average firm out there. Slightly higher
returns on equity, slightly better resilience and longevity of the business.

Back to the more general problem of valuation, you're never going to
get any two sensible people arriving at the same number.
But the good valuation methods will tend to cluster in a certain area,
and that's what should guide your investment decision.
One approach is to try a few pretty-good valuation methodologies and
apply them to the firm in each year since 1998.
(that year is chosen because the Gen Re acquisition was so transformative).
Check to see how the figures have evolved over time for each valuation
methodology, how closely clustered the various answers are, and how
the market price has tracked and diverged from an average estimate.
There is no guarantee that the average market valuation level in future
will be the same as it was in a typical year in the past, but it's
a very illustrative calculation because any reasonably good method can
be used for the following exercise: even if the valuation method is
always way too high or way too low, the historically observed average
ratio of price-to-value on that method can be used to estimate
the most likely market price today if valuations were similar to
what was seen in the past. If you think average stock valuations
will be 20% lower in future, just knock 20% off that price.

So what do I think the firm is worth?
The various values you cite in the $180k range are well thought out.
Here's another one http://www.tilsonfunds.com/BRK.pdf
But personally I'm a little more conservative and might lean towards a figure in the $160s.
For one thing, the broad US market is probably 40% overvalued right now,
so the ~$52k per share in equities might need a haircut of ~$15k.
But whether the value is $160000 or $190000 doesn't make that much
difference given the current market price of $135000.
The really nice thing about waiting is that by the time the market
price catches up to the value, the value will have grown some more,
while taking essentially zero risk of a blow-up.
I like having most of my money in a firm that will survive the next depression or world war.

Jim