Dynamic

Thanks @LongHaul And a particularly interesting video featuring Dan Ariely at the end of that post from Farnam Street (if you can sit through the first minute or so) ???

I did a search and various people on Reddit were asking about the 2017 letter and now the 2018 letter too. I suspect that with only 6 investors, IIRC, and having closed to new investors, they have strongly asked that all investors do not share the letter, and they could even have made all 6 copies different enough to identify anyone who does share. So now, it's only the SEC disclosures that would give clues to the portfolio

Seems to be about 1.7 million shares near the open, according to both Yahoo and IBKR

Hi John, thank you so much for checking my work and finding an error I'd missed! That's the beauty of sharing these things. I've now made the correction to SU and I'll edit the posts above about the changes in holdings accordingly. I may decide to implement a method of pasting the 13-Fs and parsing them to extract the info without error. I just did it for the DAL Form 4's, so I think it shouldn't be too hard to do and should reduce the likelihood of human error on my part. With that correction, my estimate of the total market value of holdings (excluding KHC) at 31 Dec 2018 is $173,342 million. This is about $585 million more than the $172,757 reported on page 12 of the Chairman's Letter in the 2018 Annual Report (and all 15 stocks there rounded to exactly my numbers after I made a couple of corrections to mine). I think it's possible that I have some more errors or that I'm unaware of or there are some pension fund holdings for companies where a SC 13/G or similar form has not been filed. It's also possible that one of the non US holdings not reported to the SEC is not at the last know size - for example Sanofi's ADRs were sold a while ago, but the last known equivalent amount of ADR is still on my sheet to reflect the majority of those that were purchased on Euronext Paris, and the year-end market value would have put it in 16th place, just outside the top 15 shown. The other potential source of error is again for foreign holdings where I haven't gone back and taken year-end currency exchange rates.

Nice. It really helps to put things in perspective

There is another Form 4 filing showing a little further buying of DAL on Monday 11th March 2019 at <$50 per share. On the assumption this keeps happening I'll try to keep the sheets updated by changing the Adjustments column on COMBINED HOLDINGS worksheet, but I won't post here about it every time. At most I might Edit this post to note new filings. If you want to see these Form 4 filings, this link (also included as a Source on COMBINED HOLDINGS column J) will show only the last 40 Ownership filings relating to Berkshire Hathaway. The Form 4 filings include the recent DAL purchases among other disclosures announced previously. Fortunately, there are no Berkshire Pension Fund holdings of DAL (see form SC13 G/A) so the only adjustment needed in column J of my COMBINED HOLDINGS worksheet is the total of purchases since the last 13-F date of record. DAL Delta Air Lines Calculated Date(US) ShrsAdded Avg Cost Cost of lot Holding % held 12/31/18 65,535,000 9.65% 03/06/19 1,202,961 $49.6514 $59,728,697.80 66,737,961 9.83% 03/07/19 1,817,318 $49.2722 $89,543,255.96 68,555,279 10.10% 03/07/19 3,600 $49.7716 $179,177.76 68,558,879 10.10% 03/08/19 2,345,077 $49.3728 $115,783,017.71 70,903,956 10.44% 03/11/19 6,500 $49.9382 $324,598.30 70,910,456 10.44% TOTALS 5,375,456 $49.4021 $265,558,747.52 70,910,456 10.44%

Thanks, gfp. That's very enlightening, and it's good to know that the fact-checking was as thorough as we'd expect.

Berkshire Hathaway has filed Form 3 and Form 4 with the SEC in respect of Delta Air Lines, Inc (DAL) common stock, first disclosing that they had become a 10% owner then giving details of their recent purchases. This 10%+ ownership changes their reporting requirements, I believe, and would, I believe require them to report buy and sell transactions within a fairly short period of such transactions taking place. I know they have in the past applied for special dispensation to delay publication of their reporting, but I suspect the SEC is not inclined to grant this while they build their position. Between 6th and 8th March they purchased an additional 5,368,956 shares of DAL and spent about $265,234,000 on these shares, giving a weighted average price of $49.4014 per share for the new purchases. Between 13-F filings, I've now added the 5,368,956 shares to my sheets as an adjustment on COMBINED HOLDINGS under "Assumed non-US holdings & adjustments" At 31 Jan 2019, DAL had 678,950,098 shares outstanding, so BRK's new holding of 70,903,956 represents a 10.44% stake in the company.

And here's today's press release. Below are a couple of excerpts:

I compared the top 15 holdings shown on page 12 of the Annual Report to my assumed holdings and found that there were three errors, while everything else matched, so I have modified my post showing the changes from 2018Q3's portfolio. Not only did the shareholdings match, but when hard-coding the closing prices at 31st Dec 2018 the value of each holding rounded to exactly the same figure as shown on p12. The errors were: 1. US Bancorp (USB) I completely missed out 4,840,500 shares with manager code 2,4,11. These are now included and the shareholding is correct. The only pension fund holdings are those with manager code 4 and I correctly exclude these. 2. Southwest Airlines (LUV) I had slightly fewer shares because I incorrectly attributed shares held by Medical Protective Company on p9 of 15 in the SC 13G/A filing to Pension Funds. These 8,722 shares are now correctly attributed to shareholders of Berkshire, while all remaining pension fund holdings remain deducted. 3. Charter Communications (CHTR) was showing far too few shares attributable to BRK shareholders. Without a SC 13G filing, I have simply put in the adjustment of -244,445 shares as the assumed Pension Fund holding, with a note beside it referring to p12 of the 2018 Annual Report, noting that it has changed since my assumptions based on the 2017 Annual Report. A further item to note, having compared the portfolio value based on closing prices at 31 Dec 2018 to Berkshire's reported value of shareholdings attributable to Berkshire shareholders, is that my assumed portfolio value was approx $172,296 million excluding KHC (without perfect exchange rates for assumed foreign shares), while the annual report indicated $172,757 million. This implies that it's very likely that the Sanofi holding on Euronext remains in the portfolio, though by my calculation it is outside the top 15 positions. I made it the 16th largest position at year end, equating to about $1,755 million with a currency error I didn't adjust exactly, while BYD was 17th at $1,680 million (with currency error), GM was 18th at $1,515 million, V (Visa) was 19th at $1,394 million and VRSN was 20th at $1,172 million. I also updated the sheet to include the correct Book Value attributable to Berkshire Shareholders and the year-end shares outstanding (exactly matching the reported Book Value per A share equivalent). I also entered the $3,652 million in Dividends received divided by the share count into cell $AH3. This year EPS is of course crazy, only slightly exceeding the dividends thanks to GAAP forcing the inclusion of unrealized portfolio losses in EPS, thus EPS excluding dividends is tiny. Depending on your purpose, you may wish to use Operating Earnings instead to analyse the profitability of the operating part of Berkshire and ignore the effects of the portfolio. The EPS for the investees is likely to be less distorted than for Berkshire, so the Look-Through Portfolio EPS of $5.42 per BRK.B share equivalent (in cell AH70, after excluding KHC's negative contribution this year thanks to their recent writedowns, which should not be included as it is accounted under the Equity Method) is probably pretty reasonable.

It's quite interesting viewing a chart of BRK.B in comparison to SP500TR. For example, https://finance.yahoo.com/quote/BRK-B/chart?p=BRK-B If you use a mouse scroll wheel over the graph to expand to around 15-20 years, then drag the bottom of the graph near the volume lines, you can change the starting point and see whether BRK.B or SP500TR is ahead at the present day in terms of total return (ignoring taxes on the S&P 500 dividends, which don't apply to BRK.B). Even picking the top of the market in 2007-2008, you can show either BRK.B or SP500TR winning until today. If SP500TR is winning the margin is usually small, if BRK.B is winning it's usually a little bigger but not dissimilar, perhaps showing that BRK.B spends more time being a little more undervalued than S&P500. In the Global Financial Crisis of 2008-2009, BRK.B, being involved in the Financial Sector, got its price hit essentially as hard as the SP500TR. At certain other times of market collapse, such as 2003, Berkshire was not hit as hard, so a starting point in mid 2002, say, sees BRK.B way ahead, despite not including the dot-com crash where BRK.B would easily outperform the tech heavy index. To me, I'm glad to see intrinsic value and book value of Berkshire rising at a slightly higher rate than the market price, and for my holding period since 17 July 2003 purchase at a slightly rich 1.6xBV, it has outperformed the very tough SP500TR competitor by a small but significant annual amount. It's certainly not the Berkshire of old, but to me it's a great mainstay of my portfolio with sufficient diversification and superior quality to the S&P500, and worthwhile tax advantages. Buying BRK.B near low points, seems (in my case) to increase by IRR by enough to make it a superior investment in total return as well as safety, and it's a good place to park cash and let it compound while awaiting high conviction value purchases that come along rarely. I wouldn't know how to price my purchases of SP500 to low points to enhance my returns.

I'm still European, at least for the next few weeks. ::) I bought and sold the B-shares under ticker BRYN (ETR:BRYN on Google Sheets seems to give the XETRA exchange - some of the smaller German exchanges have much lower volumes of a few hundred shares per day, and so does BRKB.MEX in Mexico City I believe) once via Interactive Brokers in August. Very unusually for me, I was day trading, starting before the US market opened. Volumes seem to be a few thousand shares a day most of the time on XETRA - perhaps 3,000 to 4,000 last time I was comparing different exchanges on Yahoo Finance. I bought 100 shares at €175.49 = €17549 consideration with commission of €17.55 or 0.1%. I sold 100 shares at €178.82 = €17882 consideration with commission of €17.88 or 0.1%. I netted €333 before commission and €297.57 after commission. I didn't have live market data and used Limit orders but I imagine the Bid/Ask spread is wider than on NYSE, perhaps closer to outside regular trading hours. Generally, the mid price converted to USD seems very close to the US mid price when both markets are open. It could potentially be useful to me to ratchet up my cost basis in future, making use of my annual tax free Capital Gains allowance - it's called "bed and breakfasting" in the UK and it lets me purchase outside regular trading hours, though the pre-market in the US is similar. The other experience I have is making BRK.B trades in GBP within my UK ISA tax free account where GBP is the only currency that can be held. These are processed via London only during NYSE trading hours, but are quoted in GBP and only available as At Market orders available to accept or reject for about 30 seconds, and they seem to closely follow real-time NYSE quotes when I've had that data.

Apple has been more than a double at times such as during Q4 especially since the initial purchases in the 90-100 dollar region. With dividends it may still be a double at the mid 170 dollar range. That's pretty good going in a little over 2 years and I myself got a 24 month double before I sold my Apple in May 2018. He still looks right to have continued buying at higher prices up to the 150s and 160s giving an average cost basis nearer 140, but less spectacularly so.

Southwest Airlines is intriguing, given the recent death of its founder. In the mould of the BNSF deal, if it were an airline I'm guess they'd divest their stakes in the other airlines and would be unlikely to hold any other airline stocks in future, just as they wouldn't be likely to hold Union Pacific stock now they own BNSF in entirety (though railroad regulation is probably more restrictive than airline regulation in this respect.

Also in the Chairman's Letter p12 he states something to the effect that he and Charlie view these businesses not as ticker symbols to be bought and sold on the news cycle but as a collection of great businesses that are hard to come by at good prices that collectively earn about 20% on their equity. Read his words, they're more finely honed than mine. The reduced tax liability on any gains would make it easier to sell these if grossly overvalued, but it's still unnecessarily interrupting the compound growth of great investments and lowering the compound return by introducing a taxable event, so you'd need a pretty great alternative investment (possibly an acquisition) to make it worthwhile. I suspect we might see a few more tax free swaps like the Duracell acquisition than sales of these long established large moat positions like AXP and KO. It's possible that some investees will make mergers that aren't to Berkshire's liking and cause some sort of changes. I wonder if some good companies in cyclical industries could find a good long term home with Berkshire at the bottom of the cycle. For the right price, Berkshire would not care much about the lumpiness of the earnings generated so long as their discounted value far exceeds the price paid.

I haven't had a chance to watch the CNBC interviews yet but from what I've read, it doesn't sound like he's explicitly saying Berkshire didn't buy more Apple at the beginning of January when it spend a little time in the $140s, I think he probably said he's not buying NOW (meaning around $172), and we don't have much firm evidence of Berkshire buying much at prices above $180 or so even in 2018Q3. I certainly recall thinking they probably added most of their Q2 buys in the dips into the $150s, even though AAPL ended Q2 much higher. In December, AAPL spent only about 10 trading days below $160 and only broke below $170 around 7th December. I was thinking about at what price I'd buy back into AAPL in January, and I decided that based on opportunities elsewhere, it would probably need to be $125-$135 for me to have enough margin of safety for a large high conviction position, which would also have bagged me an extra 10-15% or so of return if those prices had come about, compared to the $145 or so it was selling for at the time for a day or two. I think it's possible that Berkshire could have bought a little more AAPL early this quarter, though the lowest prices were only available for a short period right at the start of January, not allowing Berkshire to buy a lot, and we're probably not going to find out roughly how much (if any) until 4th May, and exactly how much on 15th May.

The "bait and switch" marketing sounds rather concerning, regardless of whether it's standard practice in the industry or not, and I could see that as a potential reason to either: (i) divest (if it's common industry practice that's essential to compete, along the lines of blood doping in professional cycling - Berkshire wouldn't want to remain in that business) or (ii) take ruthless steps to deal with those involved and take a close look at incentives to firmly discourage such reputation-harming practice. Having said that, from surface reading, alarm bells were going off while I read the article that this sounds like a bit of a non-story backed up by hearsay at the moment. Also a case of Berkshire neither confirming nor denying reports in case denial at times causes 'refusal to comment' to be considered confirmation at others. The news cycle is now so fast, that I note that "Berkshire Hathway [sic] did not immediately respond to a request for comment" is becoming a regular refrain in many articles. I was actually mildly surprised that Debbie Bosanek told reporters on 15th Feb that the tiny Apple sale last quarter was not Warren Buffett's doing.

I used to think that there was a cash drag at Berkshire, but that it would be invested eventually at good rates of return. Now, I think that there is no cash drag at times like these, because the amount of uninvested cash is remarkably close to float, an offsetting liability, albeit one that does not need to be paid back for many years as it would decline, in the worst case, at about 3% per year. So in normal times, Berkshire's net worth is approximately fully invested and Book Value tends to compound at about 9-11% per year. In times of market panic, where capital is in short supply, Berkshire then tends to spend a lot of the float-funded cash very rapidly on equities, and favourable preferred and warrants, taking advantage of float leverage to juice the returns without a risk of a margin call (as they did for almost the entirety of the 1970~1996 period when they could find many bargains in the stock market that would move the needle). Their credit rating also allows them to take on a modicum of debt if necessary to fund an extra large acquisition, at very cheap rates, while maintaining their $20bn liquidity cushion, and the firm's enormous cash flow will pay off such debt in short order. This seems appropriate when it's no longer possible to operate a deep-value portfolio of small cap bargains, and for this phase of its life, now as a mega-conglomerate, where Berkshire is hoping to moderately out-pace the SP500TR each decade, with lower risk, I'm happy with it. To be clear, 2018 was not a serious bear market ripe for massive investments. The market barely fell below the levels at which they closed 31/12/2017 (itself the end of a boom year in the markets), having been flat to falling in Q1, having risen sharply in the 2nd and 3rd quarters, then falling back close to 20% in the 4th quarter to leave the market about 4-5% below where it started 2017 (and 4-5% lower than a pretty high level, is still high). Prices of a lot of things got pretty expensive in Q3, and while a few names went on sale at prices that we small investors could jump at, most large caps were not deeply undervalued for very long at all, so Berkshire had very little time to accumulate significant positions at compelling prices, unlike some private investors. If the S&P index since 2010 or so has returned rather more than the growth in the businesses, it would require quite a significant drop, to bring relative values back to 2010 levels on average. I'm thinking there's a good chance of at least a moderate recession in the next year or two, possibly even out as far as 2022, though it won't stop me buying when I think things are well priced. If market prices were to fall sufficiently in the next 2-3 years, I could see Berkshire deploying some seriously significant amounts of its float-funded cash. Berkshire needs a sustained reduction in general prices to deploy such large sums with a high margin of safety, and some point in the next 2-4 years might very well provide such an opportunity. As to buybacks, sure, I think there could a reasonable amount coming soon, probably keeping cash fairly close to float when combined with purchases of other stocks. But I don't think the time of large 'elephant' acquisitions or sweetheart financing and reputation-lending deals has passed (like BAC and GS warrants/preferreds that pay off handsomely after the Global Financial Crisis). But I think it may take a while for the private equity bubble and general stock market valuations to play out and offer decent chances of such opportunities. In the meantime, Berkshire's Intrinsic Value continues to offer more than acceptable growth, and the stock appears to remain almost permanently valued at a fair to good discount to IV, so I'm very happy to keep it as a mainstay of my portfolio and usually my default "far better than cash with a modest downside risk" option for new funds that I accumulate unless something particularly promising appears on my radar. Even if Berkshire were to drop 10-15% after I park my cash there, it's likely to be a short-lived drop, and otherwise I'm likely to earn around 10% annualised while waiting for my next high conviction opportunity (e.g. something without about 40-50% margin of safety).

The Annual Report and Chairman's Letter will be release on Berkshire's website Sat 23rd Feb 2019 at 8.00am Eastern US Time (UTC -5:00) = 13:00 UTC. CNBC has also announced that Warren will be fielding viewer questions during his chat with Becky Quick on Monday morning's CNBC Squawk Box, starting around 6:00am Eastern (11:00 UTC) and video and transcripts will be available at buffett.cnbc.com afterwards.

I believe I've now updated both sheets to reflect the current known portfolio. The links in this old post will still work.

BTW, prior to updating my Look Through spreadsheet, here is a link to the summary of position changes with percentage change and absolute change in number of shares (total number of shares is not shown). Actually, I had previously included some shares assumed to be pension shares based on difference between last year's 10-K and 13-F. The 13G confirms there are now no pension shares.

I had the AAPL position as: BRK 13-F: 247,047,762 It turns out I had it correct on the BRK 13-F sheet, but on my COMBINED HOLDING sheet I'd made some adjustments to it previously that I hadn't deleted, so I was in error. I've now copied the normal correct formula and... It now correctly says: BRK 13-F: 249,589,329 NEAM 13-F: 5,711,000 Total: 255,300,329 The 13G reports: 255,300,329 shares of Common Stock, so we have a match. Sorry for the false alarm!

There were a load of 13G filings as well as the 13-F filed on 14th Feb 2019. I looked at them all. These verified the total figures from adding NEAM and BRK filings and revealed whether pension trusts were holding some of them (which I then subtract to obtain shareholder ownership). I made a mistake earlier by forgetting to remove adjustments I'd made to the Apple totals. The Apple total now matches on 13G compared to my Combined 13-F total (BRK + NEAM), so I've removed my erroneous figures. * 15 March - I also corrected an error on SU (Suncor) holding spotted by John Hjorth in the sheets and the table below. The table below shows only the DAL (Delta Air Lines) holding at 31 Dec 2018, ignoring the additional purchases taking it over 10% in March 2019. [edit after close on 5th March 2019] I found a further three mistakes and have now corrected these figures to reflect those. The LUV error was so small it doesn't change the figures shown below. Summary of quarterly changes (see COMBINED HOLDINGS tab on spreadsheet when it is updated): BK__________ Bank of New Yor…llon Corp ___3.79% _______3,087,774 CHTR________ Charter Communi…tions Inc __-4.33% ________-307,486 COST________ Costco Wholesale Corp____ ____unch _unchanged count DAL_________ Delta Air Lines, Inc_____ ____unch _unchanged count DEO_________ Diageo P L C Spon ADR New ____unch _unchanged count DVA_________ DaVita HealthCa…tners Inc ____unch _unchanged count GHC_________ Graham Holdings Co_______ ________ ________________ GM__________ General Motors Co________ __77.78% ______19,808,285 GS__________ Goldman Sachs Group Inc__ ____unch _unchanged count HCG_________ Home Capital Gr… (CANADA) ____unch _unchanged count IBM_________ International B…ines Corp ________ ________________ JNJ_________ Johnson & Johnson________ ____unch _unchanged count JPM_________ JPMorgan Chase & Co______ __39.91% ______14,451,627 KHC_________ Kraft Heinz Co___________ ____unch _unchanged count KO__________ Coca-Cola Co_____________ ____unch _unchanged count LBTYA_______ Liberty Global …c Class A ____unch _unchanged count LBTYK_______ Liberty Global …c Class C ____unch _unchanged count LILAK_______ Liberty LiLAC Group C____ ____unch _unchanged count LSXMA_______ Liberty Sirius … Series A ____unch _unchanged count LSXMK_______ Liberty Sirius … Series C ____unch _unchanged count LUV_________ Southwest Airls Co_______ __-2.44% ______-1,200,000 MA__________ MasterCard Inc___________ ____unch _unchanged count MCO_________ Moody's Corporation______ ____unch _unchanged count MDLZ________ Mondelez Intern…ional Inc ____unch _unchanged count MON_________ Monsanto Co New__________ ________ ________________ MTB_________ M&T Bank Corp____________ ____unch _unchanged count NASDAQ:LILA_ Liberty LiLAC Group A____ ____unch _unchanged count ORCL________ Oracle Corp______________ -100.00% _____-42,691,791 PG__________ Proctor and Gamble_______ ____unch _unchanged count PNC_________ PNC Financial S…Group Inc __30.27% _______2,175,743 PSX_________ Phillips 66______________ _-22.92% ______-3,537,182 QSR_________ Restaurant Bran…ional Inc ____unch _unchanged count RHT_________ Red Hat Inc______________ _**NEW** _______4,175,792 SHE:002594__ BYD Company Limited______ ____unch _unchanged count SIRI________ Sirius XM Hldgs Inc______ ____unch _unchanged count SNY_________ Sanofi (incl Eu…v shares) ____unch _unchanged count STNE________ StoneCo Ltd._____________ _**NEW** ______14,166,748 STOR________ Store Capital Corp_______ ____unch _unchanged count SU__________ Suncor Energy Inc New____ _**NEW** ______10,758,000 SYF_________ Synchrony Financial _____ ____unch _unchanged count TEVA________ Teva Pharmaceut…td (ADR) ____unch _unchanged count TMK_________ Torchmark Corp___________ ____unch _unchanged count TRV_________ Travelers Companies Inc__ __67.32% _______2,414,703 UAL_________ United Continen…dings Inc _-15.57% ______-4,045,900 UPS_________ United Parcel S…Inc (UPS) ____unch _unchanged count USB_________ U.S. Bancorp_____________ ___6.73% _______9,226,239 USG_________ USG Corp_________________ ____unch _unchanged count V___________ Visa Inc_________________ ____unch _unchanged count VRSK________ Verisk Analytics Inc_____ _-75.84% ________-758,630 VRSN________ VeriSign Inc_____________ ____unch _unchanged count VZ__________ Verizon Communi…tions Inc ____unch _unchanged count WFC_________ Wells Fargo & Co_________ __-3.40% _____-15,806,898 [s]I'm going to make the changes to the spreadsheet soon, but for now, these are the movements in the portfolio held for the benefit of Berkshire shareholders as I have them on my private version when I'm making the edits before putting them live.[/s] The spreadsheets have now been updated with these figures.

There were a ton of 13G filings too all as at 31 December 2018 which I've just been through. Very helpful in backing out pension fund holdings and checking my totals before I publish my look through sheet update Also they express the control stake as a percentage of known outstanding stock for each holding that has a 13G The Apple holding on the 13G is higher than the total of NEAM and Berkshire 13-F filings, whereas for most holdings it is exactly the total. Previously I had assumed a bunch of the holding was pensions as it didn't show up in the annual report total. Now I'm not sure if it's only barely down far less than 1% or whether the holding attributable to shareholders shown in the annual report next week will differ. For now as the Apple 13G at the top of the list of SEC filings doesn't include any pensions I'm assuming it needs an upward adjustment from the 13F totals, keeping the total Berkshire holding almost constant. I guess with Apple reaching such high valuations in October, its possible a pension scheme decided to sell it and use the proceeds elsewhere, but it's also possible it wasn't sold after all and there's something different about the 13-F filings now.

Good spot, John. After tomorrow's close seems highly likely. For me that's on the evening of Valentine's Day so my attention will be focused on my wonderful wife, rather than on EDGAR, who would otherwise get named in divorce proceedings. :P