racemize

I would like to know a more basic question--is it true that essentially all of the float has to be in bonds or cash, e.g., throughout Berkshire's (and other insurers history)?

I like using it as a quick reference for various stocks, so I can see data very quickly. However, half of my portfolio isn't covered by them, so it just makes it a hit-or-miss resource.

From a Schloss Interview: OID: You mentioned earlier that you prefer to be - and generally remain - nearly 100% invested. As I recall, before the Crash, you were 90% invested. Walter Schloss: That's correct. Over any extended period of time, stocks generally outperform bonds. Most people who have been really successful in the securities markets say the same thing - that they're not smart enough to get into the market and out of it. So they tend to remain more or less in the market at all times. I don't know anyone who got rich owning high-grade bonds. OID: What percent invested have you been over the years on average? Walter Schloss: Very close to 100%. OID: What's the most out of the market that you've ever been? Walter Schloss: 10%. And again that was before the Crash. We try to buy good values and not worry too much about what the market is going to do. You could say, "How come you had 90% of your money in stocks if you thought the market was too high?" Our answer was that what we owned didn't seem that overpriced. But when the market went down, our stocks went down too. OID: But you were up more than the market before the crash and down less during it. Walter Schloss: That's correct. And if we'd sold out, we might have missed being up 26% for the year. We did better by being in the market and not trying to figure it out. It's the same thing today. I think the market may be vulnerable. I don't know what's going to happen. But I think I sleep better owning stocks than owning cash. But everybody's different. You should own what you're comfortable with.

It sounds like he also advocates a 2% wealth tax.

[amazonsearch]Capital in the 21st Century - Thomas Piketty[/amazonsearch] Has anyone read this? It looks interesting, but I'm on the fence about getting it. Here's a review: http://www.washingtonpost.com/opinions/capital-in-the-twenty-first-century-by-thomas-piketty/2014/03/28/ea75727a-a87a-11e3-8599-ce7295b6851c_story.html

I didn't like his examples very much, but I tended to agree that uncertainty is where a lot of the upside can be found.

I was recently sent this article by a fellow board member and finally got around to reading it--very good. I'm sorry I missed it the first time.

A simpler model, racemize: Here's a historical fact that I don't recommend as a timing tool or investment strategy, but is true nonetheless. Had an investor sold the S&P 500 index anytime it reached a price/peak earnings ratio of 19 (i.e. 19 times the highest level of earnings achieved to-date), and then simply sat in Treasury bills, possibly for years, reinvesting in stocks only when the S&P eventually declined to 14 times earnings, that investor would have captured the entire historical return enjoyed by S&P 500, with substantially lower volatility and risk exposure. I just tested this on a yearly basis from 1890, and it dramatically underperformed. As mentioned by other posters, while it is a historical fact, it is based on the characteristics of the prior returns and should not be used for prospective returns (which Hussman is basically stating in his post). Moreover, I doubt he considered the effects of taxes on his original test, which is an issue, since timing the market would introduce some amount of churn. My data set that I'm using is on a yearly basis, but again, this dramatically underperformed. And again, he didn't consider taxes, which would particularly have an impact if there were any short term gains. I could retest this on a monthly basis with some work, but even he doesn't seem to think this is a good model going forward.

If you expect US High Quality to return only 2.1%, the opportunity cost is very low. https://www.gmo.com/America/CMSAttachmentDownload.aspx?target=JUBRxi51IIDZrv0FZiZbpIB3l3m%2fhybvfBpUIeImpq8f72rdLvXx7f3CjhzC9Q%2bpmRQspgkX63FM7bVF3Dp6phYxicYC3kOpmUyxo6d05eU%3d The issue here is, even when those yields get very low (and I agree they likely are), there are lots of times where the market just keeps going up anyway. Getting out of the market is very expensive as it is not predictable (take last year, where it went up 30%). Generally, you have to be very lucky about getting the gains and missing the drops, and these expected return numbers don't have a good record of actually yielding higher returns when used. That is why I've been doing modeling of various tests that people advocate (yield spreads, CAPE, etc.).

FIRST: --2012AL Besides, look at the numbers at 2013 year end: $48 billion of cash, or 21% of equity; $29 billion in bonds, or another 13% of equity; they also have businesses that bring in almost $24 billion of new cash each year, or 11% of equity. Have you ever tried to put together a portfolio that yields 11% in cash and still can grow in value over time? I have! And have come to the conclusion that without leverage it is almost impossible! When people say “I am in Buffett’s camp”… My first reaction is to think: “No! You most definitely are not!” ::) Well, again, I don't think this is a fair comparison to an individual investor. Buffett has a huge amount of money, an insurance business that requires holding a lot of bonds and cash, and simply does not have the opportunity set that we do. Moreover, have you looked at Brooklyn Investor's comments on Buffett and market timing? Even further, we have all of these quotes from Buffett as well as his behavior while running the investment partnerships: -Warren Buffett on June 25, 1999 (Business Week) Well, my point is not to always be 100% invested, it is to not hold cash as a matter of principle if the opportunities are there. For example, if I have an opportunity that I think it going to return 15%, and I'm at my last 20%/15%/10%/5% of cash, is there a particular reason not to use it for this investment opportunity? If not, what is the reason? How do you decide when to deploy or not to deploy if you are not willing to invest when your investment threshold is met? Does that reasoning result in outperformance or not? This is generally the question I'm trying to answer. Pabrai has an increasing investment bar, based on which percentage of cash remains. However, this is also a tricky way to do things, because it depends on how often the cash gets deployed. For example, if your threshold is so high that you never deploy the last 5%, and you could have gotten say, 15% on that last 5%, then it would have been better to deploy it than to not. In other words, that strategy is very sensitive to the particular threshold being set. Let's say Pabrai has ideas that normally give him 15%, and he could put his whole portfolio in that set. (Given his long term performance (after fees) is around this number, I think this is reasonable). Let's say that he doesn't invest the last 5% unless it will return 100% per annum, by his estimation. The question then becomes--how often do those opportunities come up and how often do they actually return that high of returns? If they aren't often enough (about once every 5 years, by my testing), then the strategy doesn't work. If the threshold is lower (say 50% per annum), then they need to occur much more often (e.g., once every 2 years or so). I generally like this strategy, but it is very dependent on idea generation, and you risk the case of losing out on the good ideas you have if the even better ones don't materialize. Most other strategies are based on the macro environment. Here, I think Buffett is clear (turning back to the previous point), he does not hold cash based on the macro factors. Pabrai said this explicitly, after talking with Munger about it. I think that is what people generally mean by "I'm in Buffett's camp". This is also why I've been testing various strategies against historical data, and everything I've tested so far has failed. Most papers that claim outperformance are all "risk-adjusted" and not actual outperformance. Hussman's, posted above, is the only one I know about I need to look at further, but I'm generally dismayed about the complexity of his variables (it must not be simple if it requires CAPE, a particular moving average (how sensitive is it to that 39 week average?), and investor sentiment) as well as it not comparing results after taxes. Yes, this is a very tricky question, and one that I struggle with as well. I think this is also why having a cash generating machine (e.g., a business or, say LRE) is extremely important. It gives the cash you need at the bottom and when new investments arrive. Otherwise, we end up in the sliding scale issue discussed above with Pabrai. I tend to think that the problem can be addressed by: 1) having a cash generating machines (job, business, etc.) that can generate meaningful amount of cash for investments when needed; 2) using a sliding scale (but be very careful that you don't hold cash all the time!); or 3) accept the fact that the opportunity cost of cash is why you are fully invested and that there will be times when the best opportunities are not available (e.g., if fully invested in a market crash)--this is why I've done the ex-cash returns for various investors, and usually 3) is the answer that gives the highest returns over time (although 1 is hard to test).

Could you share your experiment/model/study work as Hussman has? Or show how his work is in error? Aligning Market Exposure With the Expected Return/Risk Profile http://www.hussmanfunds.com/wmc/wmc130506.htm http://www.hussmanfunds.com/wmc/wmc131118.htm James, thanks for posting this. It will take me a while to digest this material, and I'm not sure if I'll be able to remodel it (hopefully I can approximate it) over a longer time period. So far, what I've found is 1) no one takes into account the effect of taxes on the switching strategy, which normally obviates the excess returns (if there are any); and 2) nothing has worked out of sample. In this case particularly, the model is fairly complicated--it uses CAPE, a particular moving average (how sensitive is it to using a different moving average), and investment sentiment index. When strategies get to that level, they usually do not perform well out of sample. I reproduced a paper earlier which used the spread between S&P yield and bond yields which worked beautifully from 1970-2000 (and even worked 1970-2014), but failed horribly over a longer term (1871-2014). With regard to posting my work, I very much intend to; however, while everything I've done points to the same conclusion so far, I'm still verifying and haven't started writing the actual essay. I'm happy to explain all the experiments I've done so far, but so far, no one has seemed interested enough to bother with the in-detail explanation. Moreover, this is one reason I keep posting on the topic, I'm hoping to be proved wrong or at least verify my results against data/arguments from the other side. Edit: Here is an interesting paper on timing using SMA: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2242795

In his early days, he stayed pretty well invested. Moreover, the point of my experiments/models/studies is to verify it for myself. I have not found any reason to hold cash for its own sake (or based on general market expectations). I completely agree that one should hold cash if there are no good opportunities around. I also think the two examples you gave are not all that germane. In both cases, I suspect that they were just doing conservative investments (generally bonds) and simply could not pass up the opportunity to invest when the market lows were hit. In other words, I don't think they were investing those two portfolios for the purpose of maximizing returns. Further, while it may have worked in their favor because 2008 occurred and gave them that opportunity, I have found no evidence that it is true over long investment horizons. I guess this all comes down to this--if there is a strategy where holding cash works based on anything other than running out of opportunities, then please show that it outperforms (and I'm referring to anything, there are constant macro posts, CAPE posts, Grantham posts, Hussman posts, etc.--if those ideas can result in higher returns, then I'm all for it, I just haven't found anyone showing that they actually do. I have found lots that say that they don't.). I don't see how one can stand behind strategies without proving that returns are better than when not using them.

Who? Mr. Franklin? Mr. Vanderbilt? Mr. Sage? Mr. Baker? Mr. Mellon? Mr. Templeton? Mr. Getty? Even Mr. Buffett and Mr. Munger? No, I don’t think I am assuming anything… They have been the richest and most successful entrepreneurs in history. Period. I certainly don’t know if studying their methods and imitating them might lead to the highest returns possible over the long term… But I am confident enough those returns will be satisfactory. ;) Gio I don't think Buffett and Munger recommend holding cash for the sake of holding cash. Instead, they invest in businesses that are cheap. I haven't seen anything from Buffett that indicates he would hold more cash because CAPE was high or anything of that nature. Perhaps we are just talking past each other at this point, however.

These are extremely simple "models". All I'm doing is assuming that a certain amount of cash is held (or a variable amount of cash, based on a metric (such as CAPE or something else)) and then deployed at the bottom and it tells me what the results are. I've changed assumptions over and over again, and it pretty much always says that the higher returns come in when you are fully invested. The cost of cash is very high, generally. My issue is that you appear to just be assuming the people you quote are correct. I don't like to assume anything, really, even it appears to make sense. You are using a particular system, but how can you be sure it is the right one? How can you be sure that holding cash will give you higher returns? Perhaps you don't care about getting higher returns, but want to be conservative? I don't really know. I'm just trying to figure out what gives higher returns over the long term.

Joel, How do you know this? Do you have their results on a 5 and 10 years basis without equity hedges? If so, could you please post them and tell us where you have found them? If they weren’t better than the indices, a long/short strategy (while waiting for some great opportunities to present themselves) would be sheer folly… Why do you think Mr. Watsa would have done that, if not convinced their stocks selection could beat the indices?! ::) Gio Well, why does it matter what it is without hedges? They did hedge, and those are their results over that period. I'm sure they beat it without hedging as they are good at investing, but they did choose to hedge, so why would I give them credit as if they didn't? No, no… I didn’t explain myself correctly… There is no denying they have been waiting! While everybody else is rushing to make money, they have made none. Instead they have been waiting. You might say it is wrong to wait. You might even say it is ALWAYS wrong to wait. I don’t know… I clearly think it is not so… but I might be wrong. Anyway, what I meant is there are basically two ways of waiting: the first is to sit on cash, the second is to employ a long/short strategy, which earns some alpha but is more volatile. Either way you are supposed to “be waiting”, and to be waiting means to make no money, but neither to lose it (like the post I commented was suggesting instead!). Gio Well, my point is, 5 and 10 years are pretty long time periods. Their results are not good over those two time periods right now, in my opinion. I realize that this may be a special endpoint period for 5 and 10 years, but if they don't beat the market over 10 year periods somewhat regularly, then that is very dangerous for investors--you don't get many 10 year investing periods in your life! This is why I'm talking about measuring the percentage of outperformance for rolling 3, 5, 10 year periods so that we can be sure about what we are getting. Perhaps I'll make a spreadsheet for it.

The main difficulty is style differences impact your returns depending where you are in the cycle. Given a sufficiently long time period covering multiple cycles this effect would be reduced but still not be accurate. Why not measure at similar points in a cycle? That would take away the biggest error term in measuring performance. Vinod Well, what I'm trying to measure is whether an investor could get good returns by buying at any point in the cycle (assuming a good price in FFH's case, but this can generally be applied to asset managers). I.e., I don't want to have to force the investor to only invest at the "right" time, which is the result of measuring only cycle to cycle. If focusing on 5 and 10 year periods, this should be sufficiently long term that the manager should do well, in my opinion. If the manager only does well over specific 5 and 10 year periods (e.g., based on the cycle), then that means that the buyer has to be extremely careful about when they buy. I want the manager to do well in virtually all time periods given a long enough horizon (5-10 years, with the understanding that 5 years may be something like 85% outperformance and not 100%).

Joel, I don’t know which studies you have performed… But it seems very unlikely that in every situation possible the best percentage of cash to hold, by a strictly mathematical point of view, is always zero! How is this possible? Zero, always?! In 1929 as in 1933?! In 1937 as in 1950? In 1950 as in 1968?! In 1982 as in 2000?! In 2003 as in 2007?! In 2009 as in… today?! I am no mathematician, but as an engineer I know some of math and statistics… and your conclusion seems at least unexpected to me! Let alone the fact that in investing math and logic count only up to a certain point. It is psychology that counts the most. All the great capital allocators of the past knew this very well. And they all managed to always keep strong with a substantial cash reserve. Because that gives you calmness, and calmness gives you clear thinking, and clear thinking leads you to better decisions. Either we like it or not, we are all dealing with the future… math imo cannot get you very far. ;) Personally, I try to manage my cash reserve, and let it grow when I see frothy behavior around me, while shrinking it when some great opportunities come along. Gio Well, that's the issue with this whole discussion. The answer is 0 in every study I've done. However, from your comments, I'm not sure if you are understanding what I'm doing. Here's the issue--many people on the board post articles about CAPE or market P/Es and indicate that they are holding a lot of cash. However, I've not seen anyone show that this behavior is prudent. If it is prudent, then it should be easy to show that getting in and out of the market based on CAPE (or whatever metric being used) beats the market. I've tried over and over again to get those to work, and they simply don't. Being fully invested all of the time over time has higher overall returns than any system I've come up with, when tested over decently long periods. Now, if you pick small periods, you can show outperformance with a timing system, but if it doesn't work over all sample periods or out of samples, then I think the result is noise and cannot be relied upon. Saying it another way, what I'm testing is quite simple, if you hold cash over time, consistently, and then deploy it at the bottom, are your returns better than they would have been if you had been 100% invested the whole time? I have found no system that beats being fully invested over decently long periods. I have tested perfect hindsight bias investing 100% at the bottom (but holding cash until the bottom hits), metrics based on CAPE, differences between spreads, and a bunch of other timing strategies--none of them give higher returns than just being 100% invested, on a market level. I've also tested it for multiple investors, since picking stocks is different than the market. The way to do the test is outlined above--again, the answer is 0, unless your portfolio returns, as an investor, are much, much more volatile than the market. Taking it a step further, what you just said is "Personally, I try to manage my cash reserve, and let it grow when I see frothy behavior around me, while shrinking it when some great opportunities come along." This appears to be prudent on its face, and I would love for it to be true--I wanted it to be true when I started looking at this. The issue is, I can't show that the returns are better when cash is held consistently. Do you have your overall returns, ex-cash, calculated? I'm happy to do the test for you to see if your returns holding the cash are better than they would have been if you had just been fully invested the whole time. I strongly suspect you would have done better with no cash reserves (ignoring any business concerns and the need to have cash on hand for that).

Joel, How do you know this? Do you have their results on a 5 and 10 years basis without equity hedges? If so, could you please post them and tell us where you have found them? If they weren’t better than the indices, a long/short strategy (while waiting for some great opportunities to present themselves) would be sheer folly… Why do you think Mr. Watsa would have done that, if not convinced their stocks selection could beat the indices?! ::) Gio Well, why does it matter what it is without hedges? They did hedge, and those are their results over that period. I'm sure they beat it without hedging as they are good at investing, but they did choose to hedge, so why would I give them credit as if they didn't?

I think for individual companies, that probably works. It's hard to really know for sure, but it is very similar to just having a high bar for your investments, which is an individual thing. What I'm trying to focus on here is the situation where you still have ideas that meet your investment hurdle/criteria, but are considering holding cash for other reasons (e.g., in order to have money to put in at the bottom).

Just comparing the annualized rate of growth needed for the common stock to appreciate for the warrants to do better than the common with dividends reinvested. Earlier in the year it was only 3%, now it is 4.67%. In my view, if it compounds book value anywhere close to what it has in the past (which is significantly more than the 4.67% hurdle), then the returns are pretty decent.

Take 1990 to 2002 period. I am guessing, that the worst of the tech stock funds would have had pretty high rolling outperformance but would have had very large permanent loss of capital in the end. I understand you are trying to come up with a mathematical way but I do not think this is the solution. Vinod Ah, good point, I think if we add a requirement that this is only done for long term outperformance company/investors, then it would remove the permanent loss issue. (I would generally only want to do this for something I was interested in investing in, which would only include something that is outperforming).

This is one point I disagree with in the article. We should always evaluate performance over a complete cycle, preferably over multiple cycles. We are arguably closer to top of bull market. So the performance even very long term ending at this point would unfairly penalize Fairfax. Vinod I think the fairest way to do these comparisons is to make rolling 3, 5, 10 year periods (perhaps quarterly) and show the distribution of those periods against the benchmark. For good managers, I would expect something like: 65% outperformance for 3 years, 85% outperformance for 5 years, and 99% percent outperformance for 10 years, or something along those lines. Anyone happen to have the data for that? Edit: I'm also curious if anyone disagrees with measurement in that fashion or thinks there is a better way. It seems to me to be the fairest since it is agnostic to particular time frames and the initial years.

I think the JPM warrants are cheap as well.

Good post. I think most people have stopped believing in large market declines after the bull market of the past 5 years. When valuations decline, and it inevitably will, the results will show a different story. I like what FFH is doing. They haven't forgotten the first rule of investing: never lose money. This is a theme I've seen on the board recently. It's as if the good times will never end and markets never decline. Somehow 2008/2009 is viewed as a fairly harmless blip that we navigated through without issue. I don't remember it like that at all. I remember how the fear was so strong, not just for investors but everyone. Friends who had never invested a dime were watching the news to see how bad the stock market was doing and wondering how things would survive. Since that time I've kept cash on the sidelines. In retrospect it's been stupid because I could have juiced my returns a little more for the past five years. Yet no one knows when the next crash or downturn will happen, and the last one caught a lot of people off guard. I'd rather be patient and be able to seize opportunity than miss out. My issue with the cash argument is that I've not found any evidence that it works out. For example, Sam Neil very confidently said that he thinks most money could be made by doing nothing until the opportunity arose and then stepping in. While I think that is true on an individual stock basis, and perhaps that was what he was referring to, based on all the modeling an studying I've done, it is not true for the market as a whole (which is the context of his answer). I was completely taken aback when he said that, as I've found absolutely no evidence that it is true and he seemed to be implying that it was for the market. The opportunity cost of having cash is extremely high, on a market level basis. For an individual stock picker, the opportunities have to be compared against what you have in front of you, so the opportunity cost is what you normally get, or would get in that period, which I still believe to be high. It is much harder to model that though, so there's lots of variables. The problem with this argument is that it looks foolish. I feel like value investors, because they are so contrarian, like to hold cash and call it being "prudent". It feels good right? People get to say they held cash when the market was high and redeployed it at the bottoms, but they very rarely test to see if they would have done better just investing it with the rest of their picks the whole way through. Incidentally, this can be tested without too much effort: 1) Figure out your ex-cash returns on a per year basis: 2) model holding a percentage of cash and redeploying it at the bottom (assume you have excellent timing) 3) figure out what percentage of cash gives you the highest returns over your time period For every investor I've done this for (except for Pabrai's second and third funds), the ideal cash percentage was 0, and that was over recent time periods, including the 2008 drop. Side note: I really wanted to find out that holding cash was prudent and the right thing to do. It made sense in my head. I wanted to be contrarian and hold cash as the market went up and deploy it on the bottom. I've just not been able to confirm that it works better. I really don't like the fact that if you are 100% invested, that crash is going to hurt, and you likely won't be able to buy at the bottom, but that's what everything I've tested has said, unless your portfolio volatility is very very high (like Pabrai's was in 2008, and even with that, it was better for him to be 100% invested for his longer run, first fund).

Don’t confuse the short-term with the long-term: in the long-term FFH equity portfolio has always outperformed the indices. In the long-term they will make money, not lose it! Even if nothing truly significant happens to make them change strategy. ;) Gio I generally agree, although in the last 5 years they have not, and in the last 10 years they barely did. I don't consider 5 and 10 years to really be all that short-term.