coc

I think I’ve heard Charlie say that the reason it didn’t work is that each manager picked the idea they had done the most work on. He was illustrating a bias that we supposedly have to value most what we have invested the most in. In this case time. Do you recall where this was said? It's in the first 10 minutes of this Mohnish Pabrai talk: Thanks!

I think I’ve heard Charlie say that the reason it didn’t work is that each manager picked the idea they had done the most work on. He was illustrating a bias that we supposedly have to value most what we have invested the most in. In this case time. Do you recall where this was said?

Keep dreaming. My guess is most of that is under lock and seal. Rumors were swirling that Buffett suddenly became a champion of higher taxes because he cut a deal with the Treasury to grandfather early partners in. There are allegations of dodgy partnership situations where if a dividend were paid or if the person died there could be serious tax (if not legal) consequences. And that Buffett "solved" this problem by cutting a deal. My prediction is everyone will be shocked at what comes out once Buffett dies. I'm suspicious when the only person talking about how great someone is is themselves or their groupies. My guess (educated and somewhat informed) is there are quite a few skeletons in his closet. Of course this board will say I'm wrong, he's a saint and disregard all of this. But maybe in 3/5/10 years when he passes we can revisit my predictions. "Cut a deal" with who? And why would anyone "cut a deal" with him? And why would he taint his entire reputation, which he's spent a lifetime trying to maintain, to "cut" shady background tax deals? And what does this have to do with the price of eggs? I think the real truth is that when you're on the throne, everyone's got an opinion. You say you're somewhat informed. Go ahead and tell us who informed you and what they said. If you're going to pass on gossip at least be straight about it. The guy is going to give $100 billion to charity and we want to smear him because of some vague speculation about "cutting deals"? Go ahead and call me a groupie if you like. This stuff is ridiculous. Buffett has plenty of deep, well-known flaws but the rumor mill has nothing useful to add. As for the person who doesn't like Charlie's upholding of traditional Chinese values and calling him "racist," I think you're seeing what you want to see. Charlie hired a Chinese guy (who by all accounts loves him) to manage half his fortune and the guy goes on and on about what an amazing job China has done coming up from poverty, and how much he admires Chinese ethos and how much he admires that Chinese folks are outperforming Americans, and particularly young Mungers...and he's racist? What are you talking about? Give this a read and tell me why this 1950's racist man is so beloved by a Chinese man. http://www.distressed-debt-investing.com/2010/06/li-lus-foreword-to-poor-charlies.html

And you consider what's going on in Iraq and Syria right now to be...? Human beings are capable of a wide range of things. Just because we have trended towards some sort of better morality doesn't mean we have a progressive destiny. The future isn't written yet. We're still capable of what we used to be capable of. https://www.farnamstreetblog.com/2016/03/karl-popper-mistake-of-historicism/

Thanks for your thoughts guys. After re-visiting some past information and thinking about what you said, I think you're right and I was wrong about the float. Following is from 2007: So wrong there for me. As for operating earnings, I get where you're coming from but I'm still not sure I have it straight in my head. I suspect somewhere if the railroad, utility, dealership, industrial businesses, candy shop and so on were aggregated and publicly traded, they'd be fairly valued in the 10-12x pre-tax range. In order for the earnings of the group to grow faster than they would independently -- 15%/year for example -- Berkshire HQ must allocate a lot of capital to add new businesses to them. So you can't value Berkshire's cash at 100c on the dollar and value its operating earnings at a premium without double-counting, no?

Can you explain how Berkshire would be valued at $480-$680 billon? Thanks. Using two of the three pillars of IV Portfolio=$160k; +Earnings=$12k x 10 = $280K per A share (30% above today) +Earnings=$12k x 12 = $304k per A share (50% above today) Actually the multiple deserves to be much higher for a business that has grown earnings at a 23.4% clip for 50 years and this decade looks on pace for 15%+. In my mind, for the earnings multiple, comparables are premier conglomerates, DHR, ITW; They are selling at 18x. Berkshire is better in important ways than these two. (I am intimately familiar with DHR, worked there for a meaningful length of time) . Some more IV to chew on (few want to go there); a factor of >1.0 for the third/qualitative pillar of IV needs to be included based on the historically disproportionate retention of earnings over the past 7 years, but we will let time tell if that was deserved or not. 1 dollar is being turned into 1+ dollars as we speak. So as to not double count, I think a higher multiple of earnings will cover that. (higher than what I used above). I suspect that it will all show through the continuing earnings growth. Willing to wait for that. Thanks for your explanation. A few follow-ups. (1) I would be careful not to equate "Berkshire is trading at a 30-50% discount to intrinsic value" with "intrinsic value is 30-50% higher than today's price" -- they are not the same. The first statement implies, as I said, that Berkshire is worth $480-$680 billion. The second statement implies (and I think is what you meant) that Berkshire is worth $440 billion - $510 billion, a far more reasonable range. (In my opinion.) Your second statement implies a 22-33% discount to intrinsic value. (2) I have seen this "two column" analysis before but one part of it does not make sense to me: Why is that $260 billion investment portfolio considered all belonging to shareholders? It is encumbered in two important ways: Deferred taxes on gains in the equity portfolio ($25 billion) and, more importantly, $88 billion belongs to policyholders in the form of float. You would correctly argue that both of these will be paid at some indeterminate future time, and thus shouldn't be counted 100% against the investments - and I would agree. That means the $113 billion of liabilities are worth somewhere between $0 and $113 billion. I'm not sure what the final figure should be, but I know it should be something. If you disagree, I would ask you: Would you rather own $260 billion of cash and securities with no liabilities against them or with $113 billion of liabilities against them? Yet I never see this deducted in the two-column analysis. Deducting $40-60 billion would lower your intrinsic value by $25K-35K per share. (3) Berkshire's operating earnings figure is pre-tax, I think the multiples you're citing for DHR and ITW are after-tax. They both also have much lower tax rates than Berkshire, which hurts the apples-to-applies comparison. (Pre-tax earnings at ITW and DHR are more valuable in that sense.) Berkshire has some great businesses in there for sure, but remember that to generate that 15% per annum growth, he's had to invest a tremendous amount of capital -- it's not happening organically. To double that earnings stream again will require a massive investment. So I would find it hard to swallow an argument that values those businesses at much more than 12x pre-tax earnings, which is roughly 19x after-tax earnings. Most of them are earn good returns, but aren't growing a whole lot anymore. (4) Regarding use of retained earnings, that could add a plus factor for sure, but I would argue that's partially captured already if you value Berkshire's businesses at a 12x multiple, moreso if you go higher. Looking at the metrics of the PCP deal for example, where Buffett had to pay 25x earnings or something in that neighborhood, tells me that BRK's operating earnings growth must slow pretty dramatically from 15%. Thanks for the discussion, really helpful.

Precisely because of the problem with starting/ending dates, it makes sense for you to test your formula over a longer period (than just 1995-2015). That should be done before you speculate on future returns. Yes the indexes may be flawed or inconsistent. That does make your exercise difficult. I hope you can see that why it's not easy to draw conclusions based on macro data. Macro data can often be bad data, because it's simply too tough to add up the efforts by hundreds of millions of people and compare them over many decades. In contrast, company specific numbers are infinitely better, even though they can be flawed too. Hmm, this may be a source of misunderstanding. It's not a formula and I'm not looking to get "the answer" on future returns. I'm trying to work out the parameters. Buying an index fund means buying the underlying companies, plus or minus whatever comes and goes from the index while you own it. So analyzing the companies in aggregate is not very different from analyzing any one company. It'll still come down to what's being earning relative to price paid and how the market decides to value that over time. Sales, margins, valuation, dividends will almost definition get you there, plus or minus something for buybacks and options, other smaller things. Of the relevant variables, valuation and profitability will drive returns over shorter periods, but those things tend to do some mean-reverting over time. (How much is up for question.) Check out Warren Buffett's pieces on equity returns from 1999 and 2001 -- he gets at similar points. I'm more or less trying to recreate that today.

Hmm, make sure you read all the way through the posts. We've acknowledged many times over the correlation between earnings and stock prices. It has to do with valuations. That's why I was proposing to hold "all else equal" for the sake of discussion. This is not wasted time at all. If we can establish that in order for returns to be stronger than 5% (or thereabouts) that the valuation of the market would have to go up from the 18-19x it's at now, we can think about what to expect. This has implications for pension fund returns, index fund returns, and so on. Calling the correlation between earnings and prices a "myth" is incredibly short-sighted. They are correlated about 100% over a long enough time period. But valuation can move around very unpredictably in given any start and end point. JBTC, I haven't had a chance to do that, but I suspect it comes out close. I'd love to find out though. (I don't have the data.) I get a little skeptical of the 10% number that gets thrown around because it depends highly on starting and ending points. Also, the equivalent of an S&P 500 index or broad-market index didn't exist for the entire period. The best proxy is the DJIA, which, as you all know, is a pretty flawed look at market returns.

Can you explain how Berkshire would be valued at $480-$680 billon? Thanks.

Thanks for the responses, but you guys have to produce some data to back this stuff up, and no one has yet answered any of my important questions. (1) GDP includes net exports. If something is produced in the US and sold in Brazil, that's in our GDP. The only thing missing would be things US companies produce and sell abroad. If Apple makes an iPhone in China in sells it in China. That type of sale would, however, be included in Gross National Product -- which takes in foreign sales by US entities. When you look look at the figures, GDP and GNP are pretty nearly identical in for the United States. (The difference is less than 1%.) So this is not alone a large enough factor to make a difference. (2) Liberty, you make a fair point as to why margins have increased, and I think you're at least partially right. But will they go from 10% to 12% or higher? And will margins in software and healthcare be mean-reverting over time? They haven't been yet (as a sector) but I don't consider it impossible at all. This is an unknowable. However, even if I'm wrong and margins stay this high, all I need to establish is that they're not going higher on a sustained basis for my thesis on market returns to be correct. (3) If there is meaningful sales growth included in GDP but outside of the indexes, as with startups, then my argument would be even stronger. S&P 500 sales growth would be lagging GDP growth, and returns would be even weaker. But I suspect this is not a large factor: There have been startups outside the indexes forever and yet sales growth has still managed to be in the 3-4% range, in line with I'm outlining for the future. (4) writser, I'm sorry but I don't see the relevance of your dividend experiment here. For the US to be Country R, it's return on equity would have to go to 3% from the double-digits figure that we enjoy today and have throughout our history. If we turn into Japan, that is perhaps possible, but I wouldn't consider it likely. In that case, dividends yield would go to zero and profits would grow at 3% less frictional costs. Again, that'd be a pretty new state of affairs in the US and I wouldn't consider it at all likely to happen. Country D, as you've described it, is essentially impossible. How would GDP grow at 3% per annum with no reinvested earnings? As far as it's been so far in economic history, a country must invest in its productive resources to grow. Very few corporations and no economies can get away with zero reinvestment. Thanks all.

Let's just get this out of the way -- I'm well aware of the disconnect between GDP and returns in certain periods. As I've mentioned, there are two reasons why they diverge: Profit margins and valuations. You could have aggregate market revenues double, but if profit margins are cut in half and valuations are cut in half, the market will not perform very well. And that's what you observe in these studies you're mentioning. GDP does well but the country's businesses are not profitable or there is a very low valuation placed on that profitability. Or vice versa. Thus bringing me to my point: Unless US companies become more profitable relative to sales or the valuation on them continues to increase, it seems highly likely that we're in for weak returns in aggregate. The only valid issue I see raised is that GDP growth rates and the sales growth rate of the S&P 500 or another index would be very different. But how could it be that the aggregate sales of an index like the S&P 500 would diverge from the growth rate of GDP over a long period? Let's say that GDP is growing at 3% and the revenues of the S&P were growing at 6% per annum. Eventually, the S&P would overtake GDP, wouldn't it? Which is, obviously, impossible. So there must be some relationship between the two. From 1995-2015, the S&P 500's earnings grew at about 5.3% per annum, a period in which profit margins went from from the 6% neighborhood to the 10% neighborhood. That's 2.5% of annual return solely attributable to expansion of earnings, leaving sales growth at around 3%. Also, valuations went from less 13 times earnings to over 19, adding another 1.7% per annum. Adding those together: 2.5% (margin expansion) + 3% (sales growth) + 2.1% (valuation expansion) = 7.6%. If you look at this chart, you'll see that's almost exactly what the market delivered in that twenty year period plus or minus 10-20bps. https://www.google.com/finance?chdnp=0&chdd=1&chds=1&chdv=1&chvs=maximized&chdeh=0&chfdeh=0&chdet=1419627600000&chddm=1990972&chls=IntervalBasedLine&q=INDEXSP:.INX&ntsp=0&ei=Rf_gVqicDpC7e6PLu1A The average dividend yield being paid in that period was less about 1.8%, so total return was just short of 10% per annum. Which that brings me back to me point: If we won't be able to rely on margins expanding or valuations expanding, we're left (at best) with sales growth and dividends to drive aggregate returns. And if valuations and/or profit margins contract, that will make it all the worse. Of course, it's possible that corporations will begin to earn 12% on sales and be valued at 30x earnings, but that's a speculation I can't make and I suspect would not be very sustainable. writser, as for your comment on beating the market, again, not what I'm trying to get at here. I'm trying to figure out what the market has in store for the next 10 or so years from today's price, which has interesting implications. This thread is about the bull market in stocks. Thanks.

I won't belabor my points, but two responses. (Although I agree with some of your general points.) (1) The fact that the math suggest the biggest risk is missing it is because you're cherry-picking a successful company in hindsight. If you applied the "twice book value is just as good as once book value" mentality in general, on a go-forward basis, I think you'd find you wished you had more margin of safety in lots of situations that didn't work out as well as Berkshire. (2) The difference between 5.6% and 7% per annum over 50 years on a $50,000 investment is having $1.5 million versus $760,000. I would argue ending up with twice as much money is worth figuring out how to achieve. Anyways, good discussion, thanks for your points.

writser, Your last question we can leave for another discussion - not really worried about that here. As for the other two points. (1) This is why I held "all else equal" - profit margins and valuation. And for an economy/market like the US, I suspect that the correlation between S&P 500 revenue growth and GDP growth is very high, but I'm willing to be shown contrary data. To argue that my assumption is wrong about GDP growth matching market growth, you'll have to establish either that (a) valuations will rise relative to earnings (b) profit margins will rise © revenue growth in the indexes will be meaningfully higher than GDP growth or (d) that I'm missing a material factor. I see (a) and (b) as possible but not sustainably and am uncertain what would cause ©. I would love to hear about (d) - one possibility that's been raised is buybacks. (2) Regarding dividends, there are two sources of return when you buy an index. The first is the growth in the price of the index, which I would argue is a function of the two sources above -- growth in overall revenue and profit margins on revenue. If buybacks are large enough relative to stock issuance that would be an additional factor. Corporations, in aggregate, must retain earnings to generate that growth. So the 3% sales growth would come from retained earnings, as you mentioned. The second source of return are dividends. I'm not "slapping" a 2% dividend -- that's roughly what it seems corporations are paying out relative to price right now. If dividends grow in line with earnings, and prices grow in line with dividends/earnings, then the dividend yield should be a constant (measured from today's price). Obviously, you can get a higher dividend yield if you are able to buy the index at a lower price. Which leads me back to my original question. How would one get a return meaningfully higher than 5% per annum from today's price if the revenues of American companies in the index grow at 3% and you don't get a tailwind from valuation or profit margins. One possibility is inflation -- but then you'd not be getting any more real return.

Plus net buybacks? What does that actually amount to, net of stock options & restricted stock issued? It's possible that's a bit of a plus factor, but on a net basis I'm not sure that it adds much. If someone has the figures, I'd be interested.

The above reminds me of Richard Feyman's quote about fooling yourself and remembering that you are the easiest person to fool. I don't think it's himself he's trying to fool...just my opinion.

I know what you're trying to get at, but I've done a poor job so far making two important points: (1) The law of large numbers has forged an anchor for a business like Berkshire (and many others). In your scenario, Berkshire would be worth $36 trillion in 2065. That's a multiple of the size of the current US economy. Do you consider that likely? If not, I think we need to be a lot more sensitive to price paid because Berkshire simply can't compound quickly anymore. The numbers are too big. Thus, I don't think it's fair to negate the price-sensitive argument by simply pushing out the timeframe. (2) By assuming the future is inevitably bright, and I think too bright, you are still are not considering the concept of margin of safety. The point about price sensitivity is that the future is filled with a lot of sharks. Going from $350 billion to $36 trillion in one adult lifetime is something I'd be careful to assume.

What I find interesting is that if profit margins have peaked, and they are sitting around all-time highs, then future market performance will be a function of GDP growth and valuation. If we have nominal GDP growth of 3% and a constant valuation relative to today, do we get more than 3% + 2% dividend yield over the next 10 or 15 years? Genuinely interested in counter-arguments.

The reason I think this is wrong is that you're looking at a historical process in hindsight. There is a huge, huge difference looking forward: History looks clear but it's a fog going forward. At any point, Buffett could have died, the insurance companies could have had a major fraud, the markets could have done something wacky that would have made it hard to redeploy capital, Buffett could have made a mistake, and on and on. If you had paid 4x book value, there were a number of things that could have wiped out a lot of your investment. This is whole reason for the margin of safety principle, and why Buffett doesn't encourage people to think the way you're saying. In hindsight, yes, it was clearly the right decision to almost "pay any price" for Buffett up until about 1998. But in the moment, that would have been a mistake. Buffett only goes as far as to say "A dollar of retained earnings in the hands of Sam Walton was worth far more than in the hands of the then-managers of Sears Roebuck." So you're right, but only to an extent. To give you a concrete counter-example, many investors in Fairfax Financial did exactly what you're talking about in the late 1990's -- they saw an extremely valuable and successful team compounding earnings like crazy and bid up the stock to 3-5x book value, far above any reasonable intrinsic value in the "point in time" sense. The argument went, I'm sure that it was worth it to "pay any price" for a team as good as Prem Watsa & Co., because even if they slowed down a bit, they were still pretty small and could continue compounding fast. Then the company made a bunch of unforced errors. The stock went from 3-5x book to half of book value, and today sits at a modest premium. You would have had to withstand a major, emotion-testing reversal of business fortune (the price didn't just drop due to irrationality, although maybe it went too low at the bottom ticks) -- and even now, your return would be pretty mediocre for the period. Which is all to say, while no reasonable person could disagree that you'd ascribe more value to the right jockeys and right assets than otherwise, the future is foggy and the best defense I can see is to use tools that leave me with some margin for error. Thanks for the discussion - again hope to be helpful. I agree when we talk about 5 times or 10 times. This is dangerous stuff because one is paying for quite a few years "up front" so to speak. And as you point out jockeys are mortal, and even the best assets are not invulnerable. But within the usual range? 1.1, 1.3, 1.8 times BV? I think this is a painfully false precision when one has studied and developed conviction in the three elements. The variation between the top of the range and the bottom is not an accurate reflection of risk of death, fraud etc. For the long term investor the margin of safety is nearly the same at 1 times book as 2 times book. Buffett has encouraged this nitpicking. And of course he has his reasons. As you point out many platforms, people and assets have over the years undeservingly been ascribed high multiples. But on the flip side the mathematics of compound interest are not properly appreciated. Neither broadly in society nor even among the value investing community. It's always mentioned sort of as an aside. Of course everyone can do the arithmetic. But it has a brutal power and we should, like Einstein, honor it as the eighth wonder of the world. Instead we protect ourselves with various liquidation-type valuations as if our investment outcomes were going to be measured after a year or two and plus or minus 20% was the defining factor. But for all the young people starting out the real risk towering over all others is not participating. I think you're getting at the "art" of the whole thing, which is, at what point am I not nitpicking, but overpaying? 1.5x book? 2x book? 3x book? At some point, we have to bring everything into the realm of numbers. And it will differ from business to business and even at different parts of the lifetime of the same business. 2x Book value for Berkshire versus 1x will make an enormous difference to your return 15 years from now, because the company's size dictates that it doesn't compound as fast any more. (Let's say BV is $450 per B share in 15 years. If it trades at book value then, buying at book value today gives you a 10% per annum return. Buying at twice book value gives you a 5% per annum return. Big difference!) But this was less true 30 years ago. Where we wholeheartedly agree is that, if you've got the right opportunity, trying to time the multiple down to the last tenth of a point is a mistake. For sure.

The reason I think this is wrong is that you're looking at a historical process in hindsight. There is a huge, huge difference looking forward: History looks clear but it's a fog going forward. At any point, Buffett could have died, the insurance companies could have had a major fraud, the markets could have done something wacky that would have made it hard to redeploy capital, Buffett could have made a mistake, and on and on. If you had paid 4x book value, there were a number of things that could have wiped out a lot of your investment. This is whole reason for the margin of safety principle, and why Buffett doesn't encourage people to think the way you're saying. In hindsight, yes, it was clearly the right decision to almost "pay any price" for Buffett up until about 1998. But in the moment, that would have been a mistake. Buffett only goes as far as to say "A dollar of retained earnings in the hands of Sam Walton was worth far more than in the hands of the then-managers of Sears Roebuck." So you're right, but only to an extent. To give you a concrete counter-example, many investors in Fairfax Financial did exactly what you're talking about in the late 1990's -- they saw an extremely valuable and successful team compounding earnings like crazy and bid up the stock to 3-5x book value, far above any reasonable intrinsic value in the "point in time" sense. The argument went, I'm sure that it was worth it to "pay any price" for a team as good as Prem Watsa & Co., because even if they slowed down a bit, they were still pretty small and could continue compounding fast. Then the company made a bunch of unforced errors. The stock went from 3-5x book to half of book value, and today sits at a modest premium. You would have had to withstand a major, emotion-testing reversal of business fortune (the price didn't just drop due to irrationality, although maybe it went too low at the bottom ticks) -- and even now, your return would be pretty mediocre for the period. Which is all to say, while no reasonable person could disagree that you'd ascribe more value to the right jockeys and right assets than otherwise, the future is foggy and the best defense I can see is to use tools that leave me with some margin for error. Thanks for the discussion - again hope to be helpful.

I believe Buffett sees the concept of intrinsic value as what's "there now" -- in other words, almost all of Berkshire's value in 1965 was "on the come." Buffett built that value through an interesting program of capital allocation, but it wasn't there in 1965. And if it isn't "there," I don't think Buffett includes it in his calculation of intrinsic value. If he had sold BRK in 1965, the whole company, it would probably have fetched book value +/- some small amount. With that said, he has mentioned in past letters that there are three buckets of value at Berkshire. (1) Investments (2) Operating earnings (3) Value of reallocated earnings He calls (3) a "qualitative" variable. I believe what he means is -- how much do you want to overpay versus today's intrinsic value to get the benefit of his capital allocation? Charlie also used to state the intrinsic value of Wesco (using the words "intrinsic value") as its liquidating value plus a portion of the value of its deferred taxes on unrealized stock gains (because they're paid back at some indefinite future time). He didn't add a premium for his skills that would add value "on the come." If Wesco stock was above this value, he made sure to point out that the stock was above intrinsic value. So I don't think it's incompatible to say BRK was trading about its value in 1965, and yet you'd reasonably have chosen to overpay, or even fairly pay, to get the possibility of Warren doing interesting things. Saying BRK was worth 10x book in 1965 is really not how Buffett or Munger would see it. If they had sold BRK to someone else, it wouldn't have sold for 10x book, and maybe not even book. The business wasn't worth that. The guys in charge added all that value. It wasn't at all clear in 1965 what would happen. Similarly, I don't think it's fair or smart to take Wal-Mart's market cap today, discount it by 10% per annum back to 1968, and then say the business was worth $3 billion at a time when Walton had $12 million in sales and was earning $500,000 after-tax. That's crazy and would lead to a lot of mistakes. The concept of intrinsic value should be a tool to serve your craft, not a mathematical holy writ. I use this same framework to look at most businesses, and I find it valuable to separate those two buckets (value today vs value on the come) so I am clear on what I'm buying. Hope that helps some.

Before I debate this, I want to make sure I'm clear on what you're saying (or maybe I am not), so I'll ask: Are you saying that without having worked in an industry, you cannot understand it well enough to invest in it correctly?

I think what most of you are proposing is a corollary to the idea that A stock doesn't know you own it; which is that A stock doesn't know how much research you've performed on it. You're either right or wrong. Research is a means to an end.

Very cool, thank you. I think Buffett tempered his take on large companies a bit over the years. Some of his largest successes like Coke, American Express, Petrochina and so on were large companies, and he regrets not investing in Wal-Mart, which again was quite large. Don't forget that Buffett is a learning machine - he learned as he got older.

Thank you for posting this -- really cool piece of history.

Do you think Google and Facebook have more predictability of earnings than the Burlington Railroad, Berkshire's reinsurance business, Coca-Cola, American Express, Wells Fargo, Heinz/Kraft Food, Berkshire Hathaway Energy (the utility business) and the other large components of Berkshire? Do you feel as confident that FB/GOOG will deploy incoming capital successfully as Berkshire? I am curious. My guess is that, the insurance business excepted, these major units are roughly a hundred years old on average, are in the same businesses they basically always have been in, and the reinsurance business itself is the only one like it in the world with no close second. It's one thing to be optimistic about Facebook's future profitability -- that's going to be a judgment call. And stock valuation makes the comparison even harder. But predictability of earnings?