On the subject of banks, I think they could provide an outlet for around a hundred billion of dollars more of sensible capital allocation for Berkshire if the bank holding company rules are relaxed in the next year or two and Berkshire is granted confidential treatment by the SEC, the second condition being rather unlikely.
Already, Berkshire has passively come to own slightly over 10% of Bank of America (BAC) thanks to their buybacks, yet has actively avoided exceeding 10% ownership of Wells Fargo (WFC) with whom I believe Berkshire has a closer business relationship. At the prices they apparently paid, I'd say that these represent 10%+ return opportunities with decent dividends and with buybacks to further improve returns by increasing ownership percentage over time.
Right now, with the 10% rule in place but 25% being considered as a new threshold for being deemed to be a Bank Holding Company, Berkshire has banks as 33.7% of its total equity portfolio (excluding Kraft-Heinz but including the OXY preferred), with the following holding weightings in Financials, based on my quick perusal of the list at 3rd Dec 2019 closing prices:
BAC 13.24% Bank of America
WFC 8.97% Wells Fargo
AXP 7.52% American Express (not a bank)
USB 3.75% US Bancorp
JPM 3.32% JP Morgan Chase
MCO 2.36% Moodys Corp (not a bank)
BK 1.73% Bank of New York Mellon
GS 1.70% Goldman Sachs (investment bank)
V 0.82% Visa (not a bank)
PNC 0.62% PNC Financial (regional bank)
MA 0.60% Mastercard (not a bank)
MTB 0.37% M&T Bank
TRV 0.34% Travelers Corp (not a bank P&C Insurance)
SYF 0.33% Synchrony Financial (not a bank)
GL 0.27% GlobeLife (was Torchmark - insurance)
BIT:CASS 0.05% Societa Cattolica di Assicurazione - Societa Cooperativa (not a bank Insurance)
Total Financials 45.63% ($107.2bn)
Banks only total 33.70% ($79.2bn)
With an increase to 25% bank holding company rule, there's at least scope for bank investments valued at $200 billion at current prices at about the current ratios of the existing 10% or less holdings, meaning that if purchased at current prices, $120 billion of cash could be used on purchasing stock in banks.
There is one stumbling block, however, separate from the Fed's bank holding company rules.
SEC rules require companies to report trades within 5 days if they own more than 10% of a company and these would be made public unless a confidential treatment order is granted (which is exceptionally rare these days, it seems). I could imagine Berkshire as a committed non-bank-holding-company and an entirely passive owner, might be considered for confidential treatment, but this is far from certain and perhaps is considerably lower than a 50% probability.
BAC's volume of around 57 million shares per day at $33 gives about $1.9bn traded value per day. So in 62 trading days a quarter, buying 10% of volume, Berkshire could conceivably add $12 bn or so (360 million shares) to its BAC position of almost 950 million shares in a quarter. Perhaps double that at 20% of daily volume would be around 720 million shares, $24bn. WFC might be around $5-10bn per quarter as a ball park and the other banks combined might account for another $10 bn or so. It is probably possible at the upper bound of fairly aggressive buying for Berkshire to deploy about $40-50bn a quarter into banks, if they can do so under SEC confidential treatment provisions. It's very hard to guess what the effect on prices would be if they had to report within 5 days, but they might perhaps deploy a few billion dollars (e.g. $1-5 bn but probably not much more) before their trades were disclosed publicly.