Jurgis

OT? I think that's true for science, arts, sports. I am not sure it is true for business. I am not sure Mark Zuckerberg, Sergey and Larry are somehow generational talent. I think it was more luck (being in the right place at the right time) than generational talent. Bezos maybe. And even there I'm not sure. Investing too - though perhaps a bit different. Sure there's Buffett and ???, but other than that even people considered the best 100 (or 1000) have rather mediocre long term results. Worldwide and US-wide that's true. If you look at smaller pools, it's not that difficult to be in 0.1%.

It's not surprising that for some people dollar signs is the only moral value and they can rationalize investment in anything however corrupt and despotic it is. Also not gonna be very surprising if Putin screws them and then they gonna expect to get their money back through lawsuits and sanctions of the countries which they accused to be on par with Putin's Russia.

Retirement/SS age just has to go up.

This is a rebuttal to a 2013 publication. To better understand, I would have liked to hear more about the "biases and agenda" of the authors. Submitted as complementary information: http://annals.org/aim/fullarticle/1789253/enough-enough-stop-wasting-money-vitamin-mineral-supplements https://www.aafp.org/afp/2018/0215/p226.html https://cmajnews.com/2018/06/19/trump-administration-shutters-clinical-guidelines-database-cmaj-109-5624/ The video is not a full rebuttal really. She raises a number of issues with studies that are possibly reasonable, but mostly pushes towards "we don't know", rather than "we know they are good" (which would be a real rebuttal to "we know they are not good/useful").

There is no "staying low" and I hope Stephen Hawking understood that. A civilization advanced enough for interstellar travel is very very very likely advanced enough to detect us. If you want to avoid being steamrolled by advanced civ, the good news is that we don't see any Kardashev scale III civs ( https://en.wikipedia.org/wiki/Kardashev_scale ) in our galaxy or other galaxies. We could still be steamrolled by Kardashev scale II (or scale 2.5) civ, but we don't see any of them nearby either. A scale II+ civ would very likely detect life and possibly radio signals from our planet within the radio/light/etc sphere that goes let's say from 1950s or so. So within 50-70 light years right now. Even if we don't stay low and tight beam info out, this is still limited by speed of light and very likely would only increase the chance of detection by civs that don't have interstellar travel (since civs with interstellar travel would detect us tight-beam-or-no-tight beam).

LOL. Now I feel responsible for your well-being. Drink a lot of fluids, don't travel to Russia for investing, and don't buy BRK CFDs. 8)

This is really funny. It appears I should have posted http://www.cornerofberkshireandfairfax.ca/forum/books/red-notice-bill-browder/ here instead of in http://www.cornerofberkshireandfairfax.ca/forum/investment-ideas/sbrcy-sberbank/10/ thread... 8)

Dissolving the Fermi Paradox: https://arxiv.org/abs/1806.02404 An interesting stats based paper on why there is no Fermi paradox and why we may the only civilization in our galaxy or even observable universe. Actually authors' conclusions are pretty much based on broadness of uncertainty in one of the parameters in Drake equation. But I won't spoil it fully... you have to read the paper. 8) Conclusions are a bit depressing. Especially if we manage to blow ourselves up. ::) It's a bit surprising that no one until now approached this question in the way authors did. Maybe it shows how tough it is to think differently. Should we evaluate our investments the way the authors evaluated Drake equation? I think this is somewhat related to what racemize tried to do in one of his articles and to what Damodaran does. 8)

Does anyone know how GDPR affects EU startups? Are there any exceptions for very small cos? Extended implementation deadline? Anybody's knowledgeable on this on the board? I guess similar question can be asked about US startups, since you mostly can't block EU users. But I guess the risk of adverse enforcement might be lower. Edit: from short Bing/Google-fu, there are no exceptions and there is no extended timeline. If this was enforced on everything, it would be death to startups. Of course, it's likely not going to be enforced on tiny cos though it might depend on the aggressiveness of the member-country enforcement organizations. Let me know if you know this in more depth and disagree or have further thoughts. 8)

This tool is very commonly used in dictatorships. Just read Orwell. Those who grew up behind the Iron curtain know this very well. Happens in investment discussions too... Not always consciously ...

You could try a reinforcement learning approach rather than just a supervised learning approach. The upside here is the algorithm could learn to deal with risks and optimize a portfolio. The methods discussed in the openai post TRPO and PPO are very powerful both theoretically and practically and PPO is really easy to implement. I don't know reinforcement learning in depth. I wonder if there's enough data to run RL on stock prices. Unless you do it on intraday pricing, which I don't really want to do. I think it's the same issue as with supervised learning: 10 years of daily data is only 3500 data points or so. With only 2-3 crashes in data set. But I'd have to read up on RL to see if there's a way to apply it. If/when I have time. Thanks for bringing it up. 8) The best returns come from intraday data algorithms. Not the fundemental type analysis we are all used to. The reason is these algorithms may be able to average like 10 basis points after costs (just an example your algos probably arent that good). But if your holding periods are a couple of hours or even minutes, you can make 100%+ in a year which is just not attainable with any longer horizon algorithm. Thanks for comments. You are likely right, but I have very little interest in intraday-based algos for variety of reasons. 8)

Echo chamber thread for 7 years and 1000 pages. Or at least since merkhet left.

You could try a reinforcement learning approach rather than just a supervised learning approach. The upside here is the algorithm could learn to deal with risks and optimize a portfolio. The methods discussed in the openai post TRPO and PPO are very powerful both theoretically and practically and PPO is really easy to implement. I don't know reinforcement learning in depth. I wonder if there's enough data to run RL on stock prices. Unless you do it on intraday pricing, which I don't really want to do. I think it's the same issue as with supervised learning: 10 years of daily data is only 3500 data points or so. With only 2-3 crashes in data set. But I'd have to read up on RL to see if there's a way to apply it. If/when I have time. Thanks for bringing it up. 8)

+1

Just cross validation during bull market years? I've played around with it a bit but never been comfortable enough with the algo [even worst with NNs]. I'm very scared of blowing up with these over-fitted models that have only seen rising markets... I think the main criticism against these "paper" strategies is you have 1000s of academics looking for signals and the winners publish a paper. The signals they find basically are the result of survivor bias. Do you guys have slack? Maybe its time we start a CoBF slack group So I spent a bunch of time reimplementing what these guys presumably implemented. I do not get their results. My results are pretty much at level of random guessing. It's quite possible I am not doing something the same way they did. As I said before, I'll link to their paper once it's publicly available and someone else might be able to replicate their results ... or not. I may also post or send my implementation to anyone interested after the paper is publicly available so people can shoot holes in what I did... Although I don't promise to clean up the code hugely... Right now it's a prototype-level mess. 8) The dirty secret in AI research is everyone is secretly overfitting their ANNs by by fiddling with the archtecture of the model and peaking at test set results. Only the papers with actual impressive results get published so you have a publication bias. Doesn't mean a lot of techniques don't work but they likely don't work as well as the paper would lead you to believe. This was brought up upthread. In general it is true. I don't think this is what's happening in this case though, but I'd rather not get into abstract discussions of why I don't think that's the case. OTOH, I can't really explain their results either, so who knows. Let's push out this discussion until you guys have the paper. ;)

Just cross validation during bull market years? I've played around with it a bit but never been comfortable enough with the algo [even worst with NNs]. I'm very scared of blowing up with these over-fitted models that have only seen rising markets... I think the main criticism against these "paper" strategies is you have 1000s of academics looking for signals and the winners publish a paper. The signals they find basically are the result of survivor bias. Do you guys have slack? Maybe its time we start a CoBF slack group So I spent a bunch of time reimplementing what these guys presumably implemented. I do not get their results. My results are pretty much at level of random guessing. It's quite possible I am not doing something the same way they did. As I said before, I'll link to their paper once it's publicly available and someone else might be able to replicate their results ... or not. I may also post or send my implementation to anyone interested after the paper is publicly available so people can shoot holes in what I did... Although I don't promise to clean up the code hugely... Right now it's a prototype-level mess. 8)

::)

The hard part of healthcare is that everybody knows there is not one solution. Its going to take many solutions from overtesting, to price differences in drug to a real discussion of our lifestyle and the outcomes that come from it. One of the strangest things that I have seen as I have gotten older is how much we spend on end of care and yet how many people don't get decent end of life care. I grew up near LaCrosse Wi and I remember everytime going to the doctor they always asked my Mom is she had her End of Life Directives done. It simply was a question right after can we see your ID and insurance card. It wasn't unique to me until I went to college, moved away and started my own family that I realized I didn't get those questions. https://www.npr.org/sections/money/2014/03/05/286126451/living-wills-are-the-talk-of-the-town-in-la-crosse-wis Hopefully, having these conversations with not just your health care provider but also your family leads to less issues, less costs for society and a more peaceful process. I volunteer my time at a local Hospice facility. What has struck me is the the variance of how people and families deal with this. One of the things I do is I try to help some people get their financial records in order. I've been thanked numerous times by both the people and the families and its real. Great post. 8) I'd send you $20K, but I'm not Charlie Munger.

Live and learn. Live and learn. ;D

Actually zero people on CoBF have posted their portfolios and (audited) returns. The annual return threads do not include portfolios and are mostly anonymous polls with very few people posting their returns explicitly connected to the poster. Even the ones who post their returns, don't post their portfolios. So your request is IMO out of line.

I expected better from longtime CoBF participants. Although I don't completely agree with alwaysdrawing's opinion, I am surprised about toxic and dismissive comments. You guys can be better... I hope. Peace

The dark side of trying to control overtest/overprescribe/overoperate via blunt instrument: https://www.nytimes.com/2018/06/21/well/the-strategic-lies-of-oncologists.html

I liked "The Cost Conundrum" article - though I did not send the author $20K check. 8) But based on the article I am not sure there are easy solutions. Yeah, change doctors' incentives, so they don't overtest/overprescribe/overoperate. Easier said than done. The right incentive would be "pay for great long term health results with low costs". But this is not trivial to measure and optimize for. You cannot pay the doctor 10 years after they saw the patient based on the outcome and tests/meds/operations prescribed or not prescribed. And once you try to adjust this to short term rules, either results or costs will likely suffer or there will be incentives to game the system. Yeah, someone will bring up Canada or Europe as counterexample. Maybe. I'm not sure incentives are great there. Perhaps some - most? - doctors do their best within the imperfect rules. Like he mentions Mayo Clinic culture, which is all nice and well. But it's something that's possible in couple locations driven by non-monetary incentives (culture), which is IMO almost impossible to spread across majority of doctors and hospitals. Especially doctors/hospitals that have previously run on monetary incentives of overtest/overprescribe/overoperate (sometimes somewhat subconsciously). I'm somewhat in the camp that there are not many low hanging fruit unless you socialize large swaths - which he cannot do and which would have its own issues. But maybe he has ideas that will work on large scale. Maybe he'll go for things that seem to work like Kaiser model. Good luck to him.

Two cats. We feed dry and wet cat food. Dry food as LC. Wet food Friskies one 5.5oz/156g can for 2 cats per day... so $.50 or so for 2 cats per day, so $.25 per cat per day. (I think there usually are additional discounts).

Put it all on red SPY same month calls. YOLO! 8) Seriously, IRL I'd probably do nothing. Even with special situations, there is no guarantee they'd work out in a month and they likely won't be big enough to put in $1B. The best trade would be one where you are likely make a little, but have a tail risk of losing a lot. I think this can be structured by selling out of the money calls and puts simultaneously (forgot how this is called). The thinking is that even make a little percentagewise with a high likelihood is worth quite a bit with this huge sum and blowing up with $1B is somebody else’s problem. I know ScottHall agreed with you ... and it's all for fun anyway ... but I disagree with you guys. 8) IMO, under the conditions given you really do a binary YOLO trade. I really would like bet-on-red. I.e. 50% you double, 50% your $1B margin is gone - and as you said then it's someone else's problem. ;) But yeah you could decide what is an acceptable outcome for you and bet accordingly. E.g. if you are fine with getting $100M out of it, then bet on 90% chance you get 100M, 10% you lose $1B. The problem I have with strategies that generate $10M 99% time and blow up 1% time is that ... why bother? If you risk to blow up, at least risk it for obscene amount of money, not for some 10M. :P 8) ;D I am ignoring what JRM said that you might not be able to find 50/50 or 90/10 strategies because of the money involved. Although if you went for very liquid investments, you might. And to repeat once more: this is all for fun, in reality I would very very likely do nothing.