petec

Ha ha - yes - although your deflation swaps might work in the kind of scenario that would create that outcome.

23% and 18% in FFH. Would love to be fully invested, just haven't got the ideas and don't want to overpay in what I see as an uncertain macro with significant tail risks.

Agreed, although we all seem to agree that the two will be correlated. Going back to the November 1996 example, I wonder how many clear opportunities there were then? The ensuing bubble (and it was a bubble!) took KO to 40x and bid a lot of worthless internet companies up to 8,000x eyeballs per page or whatever the trendy metric was at the time. So it's quite possible that an investor looking for good, undervalued businesses would have moved to cash. Gio's point is that CAPE suggests we are at a similar point now, where further moves up will be into bubble territory. You seem to agree using a different metric (the difficulty of finding things). I agree with both of you, because both, while they are useless timing tools, are quite good predictors of long run future returns.

None of this changes the fact that competition, either from new businesses or existing ones, should erode profit margins from record levels unless the required return on equity rises. That is not how oligopoly's and duopoly's work. Just look at companies like micron or western digital. Sometimes fiercely competitive industries evolve into not so competitive industries. This is because there used to be some weak players. And the strong players were preying on the weak ones with price wars. But as soon as the weak ones are gobbled up, there is no incentive to continue price wars. A little bit of market share might be gained, but at a huge cost. Sometimes this evolutionairy process can take decades. My reasoning is that we are seeing more and more of this. Barriers of entry for new players and the fact that everyone loses with price wars increases profit margins. Only thing that can disturb this is if some new technology comes in from the outside that is not bought out. My point is that it would be very very hard to prove (to the level that I would want to base investments on) that the same dynamics have not existed in the past. Every era has had its untouchable companies (many no longer exist) and its consolidating industries (many of which became so uncompetitive that they attracted, er, competition, because that's what high ROCE does in a free market). My feeling is that margins will fall either because demand decreases or because demand rises (prompting companies to invest and compete more). I don't think the current situation, in which demand is enough to sustain margins but not enough to incentivise investment, is sustainable over the very long term. P

+1 This is possibly my all time favourite COBF post.

None of this changes the fact that competition, either from new businesses or existing ones, should erode profit margins from record levels unless the required return on equity rises.

Agreed. I'm just not sure how likely it is in absolute terms, as opposed to relative to what others expect. I'm also a little sceptical about whether it will show up in CPI (which is what matters for the derivatives) rather than collapsing asset prices.

Realistically, I deem that to be a very unlikely scenario… A scenario in which the policies of all the central banks around the world basically fail… And after Japan, also Europe and the US slide into a prolonged deflationary environment… It is really difficult to believe such an utter failure might actually happen… The two most likely scenarios imo are: 1) Markets from now on drift almost aimlessly up and down for an extended period of time. 2) The Shiller P/E of the S&P500 reaches 29-30 (from almost 27 today) during the next few months, and then the bubble bursts. In scenario 1) I vote MKL, but FFH might do very well nonetheless. In scenario 2) I vote FFH, but MKL might do very well nonetheless. Cheers, Gio I agree an actual prolonged deflation is unlikely (although it did happen in Japan), but there's a non-negligible chance that the markets panic and offer to buy FFH's protection at a very high price. Doubling BV might be a bit much but a man can hope ;)

Agreed...if international was separately listed it would trade at >2x bv imho and it's always worth adjusting FFH's equity for that. +1! ;) And that’s basically what I have been saying for some time now… But I am not voting: MKL and FFH are both great companies led by outstanding leaders imo, and great companies led by outstanding leaders tend to surprise, and surprises by definition are not predictable… I am just content thinking those surprises will mostly be on the upside! Gio You should still vote - one of these companies has an investment with no downside that could catapult bv over 1000, and the other doesn't. So the odds can't be 50/50 ;)

Not sure. But I look at a lot of similar businesses (by type) not quality) in Latin America and they can easily trade at 2x. Plus, regardless of where peers trade, I'd happily pay 2x for a company that is able to earn a high roe from underwriting (I won't put a number on it because I can't remember the exact number!) while growing fast. And their international business does underwrite well.

Agreed...if international was separately listed it would trade at >2x bv imho and it's always worth adjusting FFH's equity for that.

Therefore, Mr. Market gets a pass for not recognizing their forthcoming unusual capital gains. Long term shareholders will get paid either way. It's worth adding the obvious point that if FFH's future returns are lower than their past returns because of the starting point of interest rates, so will everyone else's be. Their returns relative to the market, the competition, and the other alternatives available to us might not be so different. In fact they might have better relative returns than before, if the underwriting is finally sustainably reasonable.

I guess my answer wasn't complete actually. I try to buy two types of investment: stocks that I don't particularly want to hold forever that are significantly undervalued, and stocks that I'd ideally like to hold forever that are priced to give me a compounded return above the market long run average (~7% over >100y). For the first type of stock, I try to allow 5-7 years for the investment to pan out so long as my thesis stays intact. Performance relative to the benchmark is irrelevant so long as I think I can double my money in 3 years or treble it in 5-7. For the second type of stock, my holding period is basically infinite again so long as my thesis stays intact and so long as the stock does not move into clearly overvalued territory. With these stocks, the index is also irrelevant because by definition it will give the same long run return as the market (it is the market!) and so if I got my purchase right, I'll beat the index over the really long term and the only reason to stress about a 5y time horizon is because other people got rich quicker than me, which I try very hard not to care about because it really messes with your psychology. So basically, I try not to let the index come into my thinking. I try to focus on my thesis and the valuation vs. my long term return requirements, and I don't see why the prices that other people pay for other stocks should affect my decisions. I should say that I try to focus on dividend payers because that way I feel like I am getting a return even as the stock falls, and of course the yield rises as the stock falls. I find that gives me more patience than I would otherwise have. P

I'd argue: a) you hold it for as long ass you are sure it is undervalued b) you might want to sell it if you feel emotions creeping into your argument for (a) c) you might want to add to it if your argument for (a) keeps getting stronger d) you want to think very carefully about whether performance vs. the index matters to you. I hold a number of stocks that I basically plan to hold forever because I think they will provide a return that will get me to where I want to be financially in the very long term. Clearly I might be wrong but the performance of the index over a short term timeframe (and yes I do consider 5 years to be quite short) doesn't matter a huge amount to me. (I'd also ask why the index has outperformed, e.g. I could argue that with margins at records and still rising, and with QE, the index performance of recent periods isn't driven by sustainable fundamentals and therefore is not a sensible benchmark for investments that were made based on sustainable fundamentals.) Pete

Great work Gio! Have you continued adding operating earnings to your investments or is that pure investment return? And, would you sell a stake in your business? ;)

Hi Eric, The sense I have from your recent posts is that you respect the guys at FFH but haven't viewed it as particularly cheap and don't share their negative view of the world. Is that sense right, and if so may I ask what made you buy the shares? Thanks Pete

Is that so? It did nothing of the sort. I would take the opposite view: we won't know with any degree of confidence whether their insurance results have gotten better for another 5 years or more. One quarter (or year-to-date).....in what was a light cat period....reveals nothing about the risks being run and how well they are being compensated for them. Sorry don't mean to be a troll....and I'm long FFH and think they're great guys and all that, but steady on! Actually, the loss triangles have been fully satisfactory which you might interpret as some degree of confirmation that FFH's underwriting has improved. However, like you, I wouldn't mind seeing another 5 years of disciplined UW before getting too excited. SJ Yes I think the loss triangles and the accident year CR's tell a very strong story. Which isn't to suggest that the question is proved, but I think the evidence is mounting.

I totally agree but I personally feel the burden of proof lies on those who believe the latter, especially when the historical series is so strongly mean-reverting and with good reason (free market competition). My view is margins will fall when the economy finally recovers properly and a) companies invest to increase capacity, b) labour rates rise, and c) interest rates rise.

I would counter: 1. If it had changed over time then margins would have risen over time but they just seem to have stepped out of the historical range quite suddenly. James Montier at Grantham Mayo makes a good case that this is due to larger government deficits, but don't ask me to repeat the argument here ;) 2. High margins imply barriers to entry. So to justify sustainably higher margins you have to justify sustainably higher barriers to entry. How do you do that? (Comes back to my Dupont argument.) 3. To buy this argument you have to be sure that the economy hasn't always looked like the new and exciting businesses are higher margin than the old ones. I suspect it has, and that similar arguments have been made every time margins have been at a peak.

In the long term I very much doubt it. I like Frommi's line of logic and I also think about it in Dupont terms. For the margin to stay high either roe has to be high or assets turns or leverage have to be low (compared to history). I'm not sure why any of those things would happen in the long run. P

Thanks, treasurehunt. Eric, re Schilling, I actually agree totally with his premise that once private deleveraging is done the private sector will power a recovery that will make government debt look much better. My issue is with the 4 year figure, both in the sense that a lot can happen in that time and because after 4 years of deleveraging at this pace the consumer won't be at particularly low debt levels. That said, I'm not under the delusion that I can predict the economic future. I just want to protect myself from possible adverse outcomes when I'm stockpicking, because history doesn't suggest that this debt-level starting point is auspicious.

Yeah I can't say I have a lot of time for Mr. Gordon! Shilling on the other hand I do have time for and I do think there is a lot happening to drive productivity. Problem is that at least for now, productivity is boosting corporate margins not consumer demand, so companies aren't investing to create new jobs to replace the ones the productivity displaces. I'm also intrigued to know whether there has been an important mix shift in government spending. As healthcare, pensions, welfare increases, does this mean government liabilities are more inflation linked than in the past? Because if so it's going to be harder for the government to inflate people into higher tax brackets to reduce the deficit (because spending will rise in tandem). Very interested to know more about Shilling's argument that Reinhardt and Rogoff have the causality the wrong way round (debt doesn't lead to growth so much as slow growth leads to debt). I think the logic is demonstrably false but he doesn't explain his in the article. One thing that would be very helpful is a decrease in defence spending. Yadayada I've seen estimates that unfunded liabilities like healthcare etc. sum to several hundred percent of GDP if you calculate a PV but I have no idea how accurate this is.

A level of debt is sustainable. Another isn't. The exact tipping point is unknown but history gives us ranges. The '50's are only relevant in that they are part of that history. They happen to be at the low end of the leverage range and presaged a period of gdp growth at the high end of the growth range. My contention is that that is not a coincidence. Government *debt* is high but is actually about half what it was (as a % of gdp) at the end of WW2. It would be way off the scale though if you included unfunded liabilities. Private debt, however, is above the entire historical range. I'm not sure on the consumer/corporate split within private but I think it is largely consumer. Going off Prem's charts private debt/gdp just under 300%, having peaked at 260% in 1933 and 320% in 2009, and troughed at 130% in 1945. It's interesting to note how much bigger private debt is than government debt: roughly 4x. Private debt at these levels is manageable only with ultra-low rates, which have scary consequences (savers get in income, yield-seeking money is forced into risky projects and ultimately lost). There isn't an easy way out of this unless we can engineer rising wages to pay down debt. Interesting point about corporate debt being outside the US. I think it might be the other way round: US companies have much trapped cash offshore and are borrowing against that locally to fund buybacks etc. But it might distort the picture a bit.

That's a fair point if we are talking a financial crisis. But the correlation of slow and volatile GDP growth with high debt levels is unaffected. I'm talking about (potentially) years of sub-par growth, not a massive meltdown. Interesting discussion though, thanks!

Yes and no. 1952 was the trough for US total debt/gdp. Then went up for a couple of decades, flatlined for a couple, and accelerated again in the '80's until the peak in 2009. Now falling, but still very high. PW has the graph in his Fairfax presentations. NB this doesn't include the PV of future unfunded liabilities, because governments don't hold themselves accountable in GAAP terms. The discrepancy is presumably in the bond market which I believe has grown much faster than GDP as banks got disintermediated. So the lenders then are pension funds and all the other bond buyers. More importantly (for me) why worry who's lending? The data are there and aren't disputed, and the implications of brought-forward consumption on GDP comps are the same regardless.