Dynamic

I have taken a good look at the situation reported in the 10-Q now. It seems that although Float has risen to about $125 billion, the cash and equivalents have increased further to $139 billion. It seems that with regard to the portfolio, if I take my quarter end value (excluding KHC) and add the 31,081,000 shares of BAC that were added at $29.00 closing price, then subtract the -43,387,980 USG holding at $43.50 Knauf takeover price, I get a portfolio valuation of $200,597 million versus Berkshire's 10-Q $200,516 million based around the previous 13F filings. I'm usually a little bit out, but it doesn't look as though Berkshire has been buying a lot of stocks net in the quarter.

Here goes, another long post! Personally, I'm not anticipating a particularly large level of buybacks this quarter, though I wouldn't exactly be shocked if it doubled from Q1 either, but still I'm not expecting a truly enormous buyback volume that some fellow investors are hoping for. I imagine with the current IV estimate and the last reported Book Value and 2018Q3 high point of reported Book Value, we'd probably need prices in the low $190s or less to see really substantial buybacks, so perhaps compared to the old 120% of Book Value, maybe we'd be looking at somewhere below 125% of Book Value (or 125% of the highest reported Book Value) as the point where buyback eagerness would really start to ramp up. It's quite possible that later this year or early next, if compound growth continues and, if BVPS on BRK.B reaches, say $163 which is certainly quite plausible, prices below $204 or so might begin to fall into the range where some buyback eagerness might become evident. Basically, we haven't really had those sort of price/book ratios since about January/February 2016 when I personally laid my hands on all the money I could to load up on Berkshire around 123% of Book Value (one tranche costing just under 120% of 2015Q4 Book Value that hadn't been released, and another a fraction over 120%), those being prior to the 2017 Tax Cut. Even the $186 low in July 2018 was 132% of known book and 128% of the as-yet-unknown 2018Q2 Book Value. The 24th December 2018 $187.11 low was 123% of last known book value (despite a -20% bear market) and 132% of the eventual 2018Q4 Book Value that was unknown at the time and it hasn't dipped a low as $190 since that single day. I'd expect large buybacks to remain a rarity. Over time, BV may become further detached from IV, but IV is probably still running at a near-constant multiple of BVPS that might be very gradually changing over time, and would change a little more rapidly if there were significant buybacks. From re-reading the Semper Augustus 2018 letter, I'm thinking that now is not a time when Buffett is salivating over equity prices generally, though there are some sensible things to buy. Today it doesn't match the position where there was huge Berkshire overpricing of the late 1990s when it bought Gen Re with Berkshire stock trading at close to twice IV, considerably reducing the exposure to the greatly overpriced stock portfolio by buying in a bond-heavy operation. I suspect that on the whole Berkshire is currently trying to do little more than invest the cash thrown off by the operating companies and other cash inflows such as dividends, roughly keeping cash and short-term investments balanced with insurance float (which I imagine will be around $125 billion at end of Q2). It will be interesting to see whether the Occidental deal during Q2 will be additional to the typical spending on equities or will have replaced some of Berkshire's equity buying appetite. We do know they added just over 31 million shares of BAC by 17th July, accounting for $0.8-0.9 billion, but the rest of their activity is still unknown. Although this cash-to-float parity may be somewhat coincidental rather than intentional, I do feel it absolves Berkshire from having an overall cash drag to count against them, it's just that Berkshire will, when the time and price is right, have ample dry powder to take on deeply undervalued investments in stocks or whole businesses, taking advantage of non-callable float-funded leverage at just the right time. To me it's all about patience and the long game, waiting for the fat pitch and locking in a substantial increase in intrinsic value when it arrives with the use of a little float leverage too. Buffett's patience far exceeds most of ours! In the mean time, compounding Book Value at 9-11% cagr over a decade or two of low inflation and low interest rates without leverage is a great performance for such a large diversified business. Now the P/BV multiple has declined about as low as it's likely to outside extreme events, I'd expect that the future BV compounding rate ought to pretty well be reflected in the stock price compounding rate, and if that remains in the 9-11% cagr region, that's a good enough return for a low risk business in a low rate, low inflation environment, especially when the current P/BV ratio probably implies a fairly limited short-term downside risk. While I can't claim to predict the short term, I suspect that a lack of heavy buybacks in Q2 wouldn't be enough to keep Berkshire's price down once Saturday's results are out. GAAP earnings will surely be huge thanks in part to the portfolio's market value increases (excluding KHC). The share of KHC's results for Q1 and Q2 may also be included in earnings if they have caught up on reporting deadlines by Friday. And the Book Value Per Share is almost certain to be up about 3.7% since Q1 clocking a new all-time high about 2.3% above 2018Q3, with prospects for 2019Q3 looking like also showing a worthwhile increase from there. It could be that a degree of measured optimism or at least reduced pessimism about Berkshire will break out in the coming weeks and months. In the slightly longer term, headlines about underperforming or barely outperforming the S&P500 could be on the wane. It's quite a tough 10-year comparison to compare Berkshire to the depth of the 2009 Great Financial Crisis bear market (2nd March 2009 was 1435 for the SP500TR, now 6053, and 1104 for the S&P500 capital index, which is now over 3000, and BRK is up only 334% versus 386% for the total return index since that trough) and this tough comparison will ease somewhat over time for the 10-year side-by-side. This is an observation I received from someone else. Even if Berkshire is re-rated upwards a little it might yet take a bear market for Berkshire to gain a clear lead. Interestingly a chart from March 2002 to today, shows that Berkshire more or less matched the index until mid 2007 then gained a lead by December as the market fell, a lead that it never really gave back, although the index almost caught it in May 2011. I think Berkshire's success if it does retain an edge over the index in the long run, is likely to be a lumpy affair where it has periods of great opportunity where it generates substantial additional value which then gets recognised in its price, punctuating long stretches of roughly market-matching performance. And there may also be moments of market mania where the index temporarily peaks well above Berkshire. I doubt that Berkshire's price is likely to ever greatly exceed IV again like the late 90s such as June 1998, but even for someone buying then, Berkshire eventually caught up after years of lagging the index. Anyway, Saturday's will be an interesting 10-Q to read through and perhaps we'll develop more of a picture of a few facets of Berkshire's capital allocation mindset.

Thanks John, At the foot of page 1 of BAC's 10-Q they state: Berkshire's 950,000,000 shares would represent 10.206% of the outstanding common shares at July 26, 2019. On July 17, they reported the press release that there were 9,342,601,750 million shares outstanding at June 30th, so at that date, Berkshire's 950,000,000 shares had represented 10.168% of common shares outstanding at June 30th. Either way, that's 10.2% rounded to the nearest 0.1% (Motley Fool had it wrong at 10.4% and I don't think they were alone in this), so only a little over the 10% reporting threshold. Berkshire would need to sell 15,739,826 shares to fall below the last known 10% threshold if that were necessary.

Incidentally, gfp, I have checked the values held in the pension plans against the 13ga filing and indeed they check out, totalling 20,742,000 shares owned by pension funds for employees and not owned for the benefit of Berkshire shareholders. That's 47.46% of the total AAL shares reported in form 13F-HR that must be deducted. Incidentally, I have made an Adjustment to the BAC total to reflect the Form 3 filing showing an increase of 31,080,000 share of BAC to reach a grand total of 950,000,000 shares (all of which are attributable to shareholders not pension funds). All the major 'reliable' press I have seen has been getting this wrong by reporting that Berkshire added over 50 million shares since 31st March counting only the 896,167,600 on Berkshire's main 13F-HR and ignoring the 22,751,400 held via New England Asset Management (they don't read the cover page of the 13F-HR submission). The 10-K reports the previous total of 918,919,000 shares and the last 10-Q reports the correct closing market value using BAC's quarter end closing price, and the 13F filings published 14th Feb 2019 and 15th May 2019, combined, also gave 918,919,000 in total. We don't know yet whether these purchases were made before 30th June or between 1st July and 17th July, though I suspect the former, meaning that all 950,000,000 will be split across the two 13F-HR filings (Berkshire and NEAM). As BAC closed the quarter on $29.00 per share, this would be confirmed by a reported Q2 holding valued at $27.6bn ($27.550 billion rounded to nearest $0.1bn). Even one share less and the rounding would make it $27.5bn ($27,549,999,971 rounded down), surprisingly enough. Just days until we see the 10-Q now! edit on 5th August: We now know that the increase to 950,000,000 shares did take place by 30th June as $27.6bn market value was reported in the Notes to the 10-Q.

It's late here but I'll try to check tomorrow. I think the 13F total agrees with my automated parsing to avoid human error, and that's 43.7 million shares. In the adjustments column I subtract all the pension scheme assets and I've left the calculation which I entered manually from reading the 13G linked in the next column, so I'll try to check the figures tomorrow. That subtracts 20.742 million shares almost halving the total on the 13F to leave only those for the benefit of Berkshire shareholders.

Thanks for finding that gfp. I actually made a minor factual error in my previous posts. At 31st December 2018, and still at 31st March 2019, Berkshire actually owned 918,900,000 shares of BAC, a modest increase on the 900,000,000 held at 30th Sep 2018, and it was that holding that represented 9.5% of BAC's shares outstanding at year end. Of those, 896,167,600 were directly listed in Berkshire's direct 13F-HR filing and the other 22,751,400 were in New England Asset Management's "Other Manager 1 2" holdings (meaning those held for (1) Berkshire-owned (2) General Re Corporation) and all were shown in Berkshire's relevant SC 13G/A filing, the numbers in which are superceded by yesterday's Form 3. As always, the cover pages of both 13F-HR filings are important to identify the relevant holdings.

From looking at the footnotes to the filing it appears that when they purchased the 50 million extra shares, they were still below 10% and it was only when they much later learned of BAC's revised share count that they realised they were now above 10% and then had to file Form 3: If they had just purchased, they would have had to file Form 13D or 13D/A which I imagine would have been the case if they had bought in the last 3 days, as happened with DAL earlier this year, which they thought they were buying below the 10% level, and within 3 days they discovered it was above 10% due to a reduced share count being announced, and having been obliged to file 13D within 3 days, they took the opportunity to buy more prior to that filing being published and filed again to report that further buying, having increased it further to 10.44%). I don't imagine they'll make any further trades in DAL unless rule 13's threshold gets changed. I remember looking at their percentage ownership in Q4 from the 10-K and noting they were not as close to the 10% mark as they were with WFC (It was 9.5% at year end versus 9.8% for Wells Fargo). I realised then that despite the recent low prices (I had recently been put to at $26.50 per share, well below 2018Q3 prices and it had gone even lower at times) they weren't going to have been able to buy anything like another 200 million shares like they did in 2018Q3, then taking them from 700 million to 900 million, and something probably below 52 million, I think, was going to correspond to about 9.9%. They didn't buy more in Q1, so I figured maybe that was it until the buybacks eventually brought them to around 10%. As it happens they have bought more, but after the latest round of stress tests and approval of increased capital return plans, it looks like even if they do have to sell some it's probably a worthwhile purchase to have bought the extra 50 million shares, assuming they probably bought a little below $30 prior to these announcements. If they can keep the shares, even better,

Thanks gfp, I was hoping to hear your perspective as I've only read briefly about the proposals to relax the rule and don't have much feel for how restricted Berkshire truly is right now. I hope it works out like that.

There's a new Sec Form 3 filling making an early reveal of the Bank of America position increasing to 950 million shares precisely (was 900 million exactly at 31 March 2019) which exceeded 10% thanks to BAC buybacks being unexpectedly large. I suspect the Bank Holding Company 10% threshold will force them to sell some immediately unless there is any room for an exception being granted provided Berkshire doesn't but any more. The last time they asked for an exception regarding Wells Fargo WFC it was refused, forcing Berkshire to sell a little WFC almost every quarter since, but with new proposals being considered to increase the threshold to 25% perhaps it could be different this time. If any of you have any insights it could be interesting. We'll have to wait until early August for the 10-Q filling to reveal the BAC position value as at 30 June's previous close price (rounded to nearest $0.1 billion), then 14 August for the 13-F to confirm the exact number of shares. I imagine the purchase was made during Q2 rather than since its end but we won't know for sure until later. The BAC price during the early May Trump Trade war slump was quite tempting.

OK, but I wouldn't call it conservative to use such a multiple or to assume such rates will persist. 15* is definitely a very low ball conservative IV however

I think over the years, Berkshire's share price has probably been better correlated to Book Value (or perhaps highest reported Book Value) than to any other simple financial metric. As I understand it this applied before the old buyback policy was put in place (first at 110% of BV, then within about a year of that increasing to 120% of BV as they agreed a block transaction at about 120% of the last published BV). I've heard that this was considered the best metric for many years by ValueLine (I think that's the name) and also Mungofitch over on the Motley Fool Berkshire messageboard keeps some statistics over a very long period comparing prices to last-reported BV and to high-watermark BV and noting how rarely, for example, P/BV falls below 1.3 and what the typical returns are. He also has a 2½-column approach valuation measure similar to Buffett's approach which he says seems to very well track the average price over a year, with ups and downs superimposed on it. He can usually quote stats such as: That $200 was in Feb and was 1.31x the 2018 Q3 BV of $152.475 per BRK.B. Mungofitch has had to modify his 2½-part valuation method to adjust for the tax cut, but it also remains a good predictor of average price over 6-12 months. He also mentioned in the same thread in February that price-to-peak book is a better predictor than price-to-last-known-book in the 1996-2016 starting points. It will be interesting to see if the close correlation between BV and share price changes anytime soon, now that the formal buyback policy has decoupled from BV and now that Book Value per Class A share is no longer specifically mentioned in the News Release that accompanies each 10-Q or 10-K filing nor will it be included in next year's Annual Letter to Shareholders. It is easy enough to calculate and does appear on various data sources like Yahoo Finance of course. Now that some of the past history of buybacks is disclosed, perhaps some people will perhaps anchor to those price ranges instead of BV multiple. If they anchor too heavily when Berkshire's IV continues to grow, it could provide opportunities for much greater buybacks if the conservatively calculated IV grows sufficiently far above it for Berkshire to consider it among the best uses of funds to buyback very heavily. I would not be at all surprised to see BV in 2019Q3 (published in November) come in well above $160 per BRK.B, probably at least $10 above the Q1 figure in the space of 6 months, albeit perhaps only about a 6% rise from the high point at 2018Q3. That could change Mr Market's mood a little, but if the share price remains close to $200-$210, it could look like a very attractive repurchase opportunity for Berkshire. I'm pretty sure Berkshire's IV has risen by at least 6% and more likely 10% since last year when a few repurchases were made at $207 average, meaning $220-$227 would make a sensible price for only modest buybacks this August, so prices below $210 might see somewhat greater buyback volume this time, and prices below $200 would probably see more significant buybacks, and below $190 they might be heavy. Still, I'd be rather surprised to see more than $3 billion spent on buybacks in Q2, though it would be a pleasant surprise. Strategically, I'd also imagine that Berkshire's buyback policy should not be linear and predictable. Better that it reflects the universe of alternative investment opportunities anticipated and looks a little random and illogically weak at times. Looking at possible alternative valuation approaches that might influence future share prices if BV loses its correlation... GAAP EPS as a metric for Berkshire's valuation has always missed a lot of hidden earnings and is crazy now that it includes unrealized gains on stock positions, so that won't be used. Look-Through Earnings (ignoring the unrealised gains) seems reasonably good, but it's not part of the official financials. According to this article from October 2018 that I've just come across... (emphasis mine) Funnily enough, I think I came up with about $15.48 per BRK.B share of L-T earnings (portfolio look-through earnings, plus operating earnings) in late Feb. As a conservative IV estimate I then applied a 15x P/E multiple (6.67% earnings yield) to come up with $232 as a pretty darned conservative IV, and would guess that prices at 80-90% of that ($186-$209 or so) would encourage Berkshire to buy back stock, more aggressively at lower prices, I'd imagine. Applying the 17x multiple that J.P. Morgan analyst Sarah DeWitt mentioned, would be $263 as a more typical price if the $15.48 Look-Through EPS still stands, with $210 representing a 20% discount to IV, and $184 a 30% discount, as another way to look at things. $264 happens to be an average of different intrinsic valuation methods published by Semper Augustus, though I'd be a bit surprised if the share price even reaches $260 during 2019 or early 2020 even if L-T earnings is now nearer $15.80-$15.90. I think perhaps the tax cut has inflated EPS so the multiple of 17 might not tie in with typical Berkshire share price under a 21% federal tax rate. Given that BV of around $150-$160 per BRK.B is about 10x Look-Through Earnings, that also ties in well with a 10% return on equity, close to the 10.98% compound growth rate in known Book Value per share and 9.30% compound growth in stock price over the last 16 years, the difference reflecting a decline from about 1.6x BV to 1.3x BV multiple from July 2003 to July 2019. Now, as a saver still, I'd actually be happy if the stock price languishes a little longer, and I'm not banking on 20%+ per annum compounding for the next 2 years (that would be around $300 by July 2021, so perhaps something around $260-$280 is more realistic - about 12-17% 2-year cagr barring a recession. I'd also be happy for the price to languish to enable cheaper buybacks and retirement of more shares well below IV.

I agree that the 3-day reporting rule would be a major impediment to buying more. Are you mainly talking about the bank holding company regulation, where anything over 10% makes you a bank holding company and produces onerous restrictions on your banking relationships etc., as that's one I've read about plans to consult on loosening the restrictions? The need to continually trim WFC as they buy back their own stock is somewhat annoying, and the same restriction will probably limit the BAC exposure to the 900 million shares they currently own, as much more would be getting perilously close to 10% there too, forcing sales that have no economic rationale, purely a regulatory compliance rationale. It could also be a major boon for non-financials if they loosened the 3-day reporting either by changing the threshold or extending the deadline in certain circumstances. That was the regulation that required Berkshire to file a 13D after Delta (DAL) revealed that they had bought back a lot of their own stock, pushing Berkshire over 10% unexpectedly. On realising the cat was out of the bag, Berkshire decided to buy a little more DAL in the 3 day window before it became pubilc, taking their position up to 10.44% at the time, filing a second 13D a few days after the first to disclose it, but then they haven't bought or sold another share of DAL (unless it was within the last 3 days!). Berkshire certainly doesn't like to make moves in open-market securities where public disclosure of their proprietary investment activity is required so soon. A little more freedom in the reporting regulations would greatly help Berkshire in 'moving the needle', and could widen its potential pool of future US investments substantially. Various other market have much tighter rules, such as the UK, with a 3% limit, though that probably help limit losses from the Tesco plc position!

I agree. I'm optimistic that the economic forces of such inexpensive power (and still declining rapidly in cost) will soon completely undermine any efforts by the fossil fuel lobby and climate change deniers funded by them to prevent the retirement of the most polluting and least efficient fossil fuel generation at first, and almost all of it eventually. I think there need to be a few incentives to encourage such large up-front capex at this point, which will create the demand that will further lower the cost, but it's pretty clear that the cost curve for wind is on a significant decline and for solar the decline is steep almost to the point of precipitous, such that it's now falling below wind on a direct price/kWh basis and is soon likely to be cheaper even when including the cost of storage typically required to time shift supply to meet evening demand. Low interest rate financing certainly helps to fund the capex, and in Berkshire's case, the tax treatment allows these technologies to provide returns above their hurdle rate, encouraging truly huge capital investment while sustaining low power bills for end users. I suspect there may be benefits to the likes of Google, Facebook, Microsoft and Amazon, locating many of their vast data centers close to renewable energy sources, not the least being in term of energy cost savings. It's also clear from many parts of the world, including Norway with it's sparse population and the UK with its large coastline for offshore wind despite its high population density, that a low carbon, high reliability grid is not the problem it is often painted as by the fossil fuel lobby. As a UK resident able to choose my supplier (but paying for the network services indirectly through their billing), my cheapest suppliers (I switch most years) have nearly always been 100% renewable for electricity, with only natural gas (methane) for heating and cooking being fossil based. (There are suppliers working on bringing green methane gas to market too). I believe it's the case that the offshore wind industry has employed a lot of former oil and gas workers whose skills are transferable to that sector after oil prices plummeted, though many helicopter pilots who used to serve Norwegian and British oil and gas rigs are no longer employed in that sector. I suspect that by switching from fuel costs paid to various oil exporting nations to capex and maintenance costs, the costs to build and maintain green energy facilities over time will continue to boost local employment rather than paying oil workers and well owners in other countries. If the costs of renewables continue to get lower, it's likely than maintenance expense per kWh will remain similar, but the capex savings will reach the consumers by way of lower utility bills, and will presumably also boost their spending on goods and services, at least some of which are likely to be spent locally. I think there's a virtuous circle that will probably increase the rate of uptake, and it's not looking like a promising time for new investments into coal technology, though gas peaker plants may have a place for a little while longer, until battery supply completely supplants those with superior response time for frequency stabilisation and peak shaving and at lower cost too. The huge automotive demand for batteries is also an important driver for rapid cost and energy density improvements in battery storage and possibly even power inverters.

Thanks John. I wonder if that's part of the capex detailed in the slide set that was linked to in the Berkshire Hathaway Energy thread or whether it's additional. Nonetheless, it's good to see continued capex that will doubtless offer a good return and in inexpensive, clean and safe generation with only maintenance and renewal costs, not ongoing fuel costs, greatly enhancing the generation efficiency of existing sites where it makes sense to do so.

I think actually (and hopefully it's not wishful thinking), you're probably right to point out that it's very likely that capex will not decline significantly and that these planned expenditures will be increased with additional capex as plans develop further. I suspect the incentives haven't changed too drastically, despite a few of Trump's intentions to boost coal and rein in renewables incentives at the Federal level and all the 'newspeak' recently about 'freedom molecules' and so on is more about messaging to his supporters than policy changes being implemented widely. Certainly the ROE allowances are amenable to incentivizing continued capex for companies as future-oriented as BHE

I do agree that Google sometimes shoots itself in the foot with product abandonment and casts doubt on continued support for a lot of new platforms where they need external developers to believe in sufficient support to be motivated to provide content. I have no idea of the cost of sustaining the old Google Finance while they developed a new version, though I'm well aware that skilled human effort is required to fix things that break due to outside influences (e.g. data feeds changing format, stocks with corporate actions such as stock splits and divestments and much more besides), so it's not as simple as some might think to keep everything that was there running. Nonetheless, they have to be aware of the potential to lose users and the long-term cost of that too. I have noticed a number of changes to the way the Google Sheets function GOOGLEFINANCE() works over time and updates to their Help documentation about it, indicating that they've probably been making a lot of changes behind the scenes. I think they have been switching over data suppliers etc. (hence support for certain ticker formats changing, such as 123.TO becoming TSE:123 and all exchanges being able to be explicitly specified) and probably replacing a lot of old code that had probably grown and spaghettified to handle the changing data formats from various suppliers of financial data. I suspect this became problematic to maintain and is being rewritten from scratch, perhaps with a different strategic approach using modern formats that offer advantages, and a different type of strategy, especially for dealing with historical data (such as EPS and prices adjusted for splits and currency changes etc). Certain odd things I could use in Google Sheets functions for simplicity that once worked ended up broken permanently or worked only sporadically, such as converting currencies to themselves or their sub-divisions such as pennies (e.g. USDUSD or GBXGBP) so I had to modify my spreadsheets to catch these cases and manually put in 1 or 0.01 etc. instead of calling GoogleFinance. But it remains a very powerful function, and it's possible to do some pretty good stuff with it in spreadsheets. I am liking some of the things they're changing (either that or things I'm only now discovering that were changed a long time ago), and I think it may help to make things quite robust, but I'm also glad that I have been using it in my own spreadsheets to help track my portfolio rather than relying on their Google Finance website portfolio functions which have been disrupted. And there are certainly a number of things where I rely on a whole host of different sites for certain data that matter to me at least somewhat, but don't seem to be widely used (e.g. FTSE100 Total Return Index). Fortunately SP500TR is on Google Finance.

He may just be basing it on BHE's own projections. Of course, their future plans don't reflect projects that aren't confirmed or opportunities that haven't come up yet. But as of earlier this year, they expected capital expenditures to decline. It probably won't actually happen. Projects are included in the projections as they are committed to. https://www.sec.gov/Archives/edgar/data/71180/000108131619000007/ic2019.htm (several mentions but slide 22 shows it fairly clearly) Yes, their own projections are for a degree of reduction from an admittedly huge rate of capex and I remember seeing something like slide 22 somewhere in the last few months. I'm not sure of the reasons, but I'd imagine slightly reduced cost of capex (meaning they need to spend less to get the same improvements would have some impact, though it looks to be modest) and reduced tax benefits (perhaps influenced by Trump's policies with regard to CO2 reduction in comparison with Obama's and the expiry of certain schemes put in place some years ago). That slide set is interesting, and I don't think that's the place I originally read about their projections of decline in capex. It's fascinating to see such a huge discrepancy between Net Income Attributable to BHE and Cash Flows from Operations on various slides, with cash flow being consistently far higher since 2001 and also being around 4.5x interest expense. The information on how tax rate reductions are being passed on to customers is also interesting. The settlements on rate reductions seem to be around 3-4% mostly, so as expected BHE will not retain the full benefit of the tax cuts and will share it with customers via rate reductions. Also interesting is certain jurisdictions having ROE limits above which a certain fraction of additional profits are shared with customers (some places it's 50%, some it's 90% that goes to customers). I note that MAE has implemented a 1MW power / 4MWh energy storage (thus 4 hours' supply at full power) battery using Lithium Iron Phosphate tech (sounds like BYD's favoured cell chemistry but not necessarily only theirs) with inverter/transformer in 2018 for $3m as a pilot project mostly used for energy balancing purposes. That's about $750 per kWh all in for a trial site (compare capacity to 40 Tesla Model S/X 100 kWh batteries or 625 first gen Tesla Powerwalls. I imagine the cells would be somewhere in the $150-$300 per kWh range, and the other costs would be for inverter/transformer and the costs of developing and installing the site and designing/running the project. It's also interesting to see MAE's asset profile versus its power profile. First, coal, gas and nuclear have declined substantially in 18-19 years. The Wind generation and other assets are now 87% thanks to huge investment but their power contribution is 59% (on a net MW owned in operation and under-construction basis). Coal 12% and Gas 1% have a much smaller asset base, but can supply 24% and 13% of peak power capacity (with Nuclear and Other providing 4%). But of course, they also consume fuel over their lifetime, which wind doesn't, although it does consume maintenance costs and a few consumables. This is a major shift and should pay off very nicely for MAE and BHE as a whole in the long term as a reward for the enormous capital infusion (funded by retained earnings and relatively cheap debt that is non-recourse to Berkshire Hathaway Inc). Let's hope that regulators and technical innovations allow plenty of scope for more profitable capex in BHE in the future.

I think the dollar amount of capital expenditure is likely to decline from the very high rate recently, where the tax deferral advantages had been hugely beneficial, but are soon to diminish from that peak. Nonetheless, I'm sure a lot of capital will still be deployed at attractive rates by BHE, often with attractive tax benefits. It now appears that for NVE, it's the three partner companies that will run these projects and presumably benefit from a long-term contract with NVE, and presumably it is those companies that will raise the capital required. I think BHE has been looking at grid-scale battery storage for quite a while - and their BYD investment many years ago was probably in part linked to the potential of grid scale storage, which was mentioned around the time as items being closely watched, back in David Sokol's day when it was still called MidAmerican Energy. 3.5 cents per kWh with storage, seems very attractive. The costs of solar and battery storage are coming down so rapidly (even faster rate of decline than the cost of wind generation, with the cost per kWh of solar alone now roughly at parity with wind alone but declining faster) such that a lot more capacity of generation and storage can be bought for less capital expenditure than just a few years ago, and now cheaper than most other sources of baseload and peak generation, so I'm optimistic that the economics emerging from these price declines will turn these low carbon alternatives into the default no-brainer solutions in the near future, and will increasingly become economically superior to keeping coal-fired power stations going, providing environmental improvements and additional employment that will far outweigh the losses in the coal industry. The Holmesdale South Australia battery plant in a short period has reportedly made a lot of money providing grid stabilisation services with extremely rapid reaction times, for which it's financially well-rewarded by the dynamic pricing system in that jurisdiction, such that I believe it makes more of its money from those functions - rapid-response peak shaving and trough-filling and the time-arbitrage it can offer - than from sustained output functions. Further facilities in the same area would have diminishing returns, but it clearly serves a valuable function and is well rewarded.

An interesting YouTube video from Home Services Relocation Services about the misalignment of incentives that has crept into the employee relocation business and what they're doing to realign it. https://www.youtube.com/watch?v=O3e24YqEi8Q I originally found the video via Seeking Alpha, and embedded in one of Home Services' pages, but it was kind of annoying how they embedded it. They're launching "True Partner" to realign the incentives.

You're welcome - I learned a lot that is useful for my Berkshire Look Through Portfolio tracking.

Thanks for the kind words. It has continued to be useful to me too, and I greatly appreciate all the help from everyone on this thread who have helped point me to new information or to spot errors, thus helping me to improve its usefulness to me and to everyone else. Nice to be part of a virtuous circle!

In doing a little detective work in a thread about Li Lu's Himalaya Capital Management portfolio, finding out why they no longer need to file 13F I came across the SEC's 13F FAQ which clarifies a lot of the rules and provides links to a list of all the securities that are covered by rule 13F each quarter (including archives, such as 2018Q2 when Monsanto, spelled correctly in the SEC's list, was deleted having been taken over). For those interested in the minutiae, there are a few nuggets extracted from the FAQ quoted on that post that may be of interest, including the clarification of the 13F filing dates, which are 45 days after the quarter ends but if that falls on a non-business day the deadline is delayed until the next business day (e.g. not a Saturday, Sunday, Presidents' Day Public Holiday etc), so while this happens to match the expected dates of 14th Feb, 15th May, 14th Aug, 14th Nov in 2019, the next filing that will be delayed is 2020Q3, where 14th Nov is a Saturday so the filing will be on Mon 16th Nov (unless that's a Public Holiday for some reason) and we won't get back to the normal 45 days after quarter-end until 14th Feb 2022 as they'll fall on weekends until then! There are also descriptions of the confidential treatment provisions towards the end of the FAQ. Apparently when the confidentiality period ends, an amendment would be filed. For some filers, but probably not Berkshire, they mention rules about not having to disclose or net-out short positions (including short options positions such as put-writing or call-writing) but the need to disclose all long positions. Interestingly, about 186 securities containing the word INDEX are on the list of reportable 13F securities, some of which seem to be ETFs and some CALL and PUT options too. In other news, alongside automatically reading in the 13F filing data for Berkshire and the relevant entries from New England Asset Management so that I avoid so much human error in future, I believe I'm homing in on a good approach to simplifying the appearance and sort-ability of my Look-Through Portfolio. I'm trying that out in my private version, so I may well update it by about the time the next 13F drops on 14th August.

In searching EDGAR, we have: Himalaya Capital Management LLC whose last 13F-HR filing was on 14th Nov 2018 for the period ending 30th Sep 2018. It shows $10m of holdings in Baidu (5,800 shares, worth $1.326m) and AliBaba (53,000 shares, worth $8.732m). The number of shares matches the previous quarter's holdings too. There has not been a filing from them since then, which might imply that they no longer hold these securities and are now exempt from filing... or perhaps they're exempt from filing for another reason: According to the 13F FAQ from the SEC: Clearly, Himalaya is well below the $100 million threshold in Section 13(f) securities, even if they still hold the same Alibaba and Baidu stocks (which both remain on the current list of 13F securities), as their BYD holding is Hong-Kong listed and exempt from 13F. The 13F-HR filed for 2017Q4 reported $7m of stocks, 2017Q3 had nearly $8m, 2017Q2 nearly $6m, 2017Q1 nearly $118m, and 2016Q4 nearly $112m (Baidu Inc and Sina Corp - possibly this having been granted confidential treatment as it was published on 14th June 2017, the same date as for 2017Q1 which showed the same holdings) Question 28 says you must start filing for the December quarter in the year you first exceed $100m at the end of a month then continue filing form 13F and will need to continue for March, June and September even if you fall below the $100m threshold: So I imagine that once they knew their 13F would become public, they decided for proprietary advantage to fairly quickly reduce their exposure to 13F securities to well below $100m. As they had exceeded $100m for at least one month-end in 2017 (March at least), they would have to file 2017Q4, and the first three quarters of 2018, but could then cease filing, their last 13F-HR report having to be filed on 14th Nov 2018. Also Question 25 specifies the dates of filings: A few other interesting items:

I did find a site (click here) listing both Himalaya and Berkshire Hathaway Holdings of BYD on Hong Kong along with BYD's two main executives who held it in Shenzhen exchange so for completeness I will change the ticker I use in teaching Berkshire's holding. The double listing made the site corrupt the percentages but the share counts matched other sources I found so I guess it's probably accurate. Maybe it's possible to check other suspected holdings to see if Himalaya Capital is listed as a major holder.

I don't have time to study their 10-K today, John, but I'll be glad to hear any thoughts you derive from a Berkshire owner's perspective in this thread, or of course, feel free to write more in the KHC thread in the Investment Ideas sub-forum.