Financial topics

[aeden]

H we understand what adaptation is.

https://www.zerohedge.com/sites/default ... k=gzm42lpa

[aeden]

https://www.cato.org/publications/comme ... rock-solid
http://www.scmp.com/business/money/mark ... cent-fears

[Higgenbotham]

Today, computers are powerful enough today for neural networks to be
trained on infinitely more data than Warren Buffett could learn from
in his entire lifetime. Not only that, it can learn on its own by
scanning the internet.

If humans are still better at some things today, that list of things
will be reduced with time.

By 2030, multiply that power by more tens of thousands, and humans
don't stand a chance. I won't be alive to see it, but maybe you will.
Enjoy!

Buffett has real time sales and other data coming in from the over 200 companies he owns and has private access to. Nobody in the world has access to this much valuable real time data.

Scanning the internet is for blokes like me. Despite the fact that my data processing capability is orders of magnitude less than today's computers, they still can't beat me at trading in a rigged game where the exchanges have given them massive advantages. If by 2030, computers have the equivalent of a hammer that is 10,000 times more powerful when what they really need is a screwdriver, they will still be losing.

The tools computers have are great for playing self-contained rules-based games like chess that have a finite number of possibilities. Humans suck at those kinds of tasks.

By getting computers involved in tasks that they have no business being involved in and for which their involvement actually confers massive disadvantages which are disruptive to the functioning of free markets, supply chains are going to collapse and the world is going to go into a dark age. This is as certain to me as the sun will rise in the east tomorrow morning.

[aeden]

https://medium.freecodecamp.org/how-to- ... 46935d93fe

https://www.youtube.com/watch?v=G87UXIH8Lzo

http://gdxforum.com/forum/search.php?ke ... sf=msgonly

https://www.youtube.com/watch?v=Zz_YGeAsbBU

[shoshin]

Higgie, "rules based games"? I have one word for you; GO.

[John]

> Buffett has real time sales and other data coming in from the over
> 200 companies he owns and has private access to. Nobody in the
> world has access to this much valuable real time data.

> Scanning the internet is for blokes like me. Despite the fact that
> my data processing capability is orders of magnitude less than
> today's computers, they still can't beat me at trading in a rigged
> game where the exchanges have given them massive advantages. If by
> 2030, computers have the equivalent of a hammer that is 10,000
> times more powerful when what they really need is a screwdriver,
> they will still be losing.

> The tools computers have are great for playing self-contained
> rules-based games like chess that have a finite number of
> possibilities. Humans suck at those kinds of tasks.

> By getting computers involved in tasks that they have no business
> being involved in and for which their involvement actually confers
> massive disadvantages which are disruptive to the functioning of
> free markets, supply chains are going to collapse and the world is
> going to go into a dark age. This is as certain to me as the sun
> will rise in the east tomorrow morning.

Really, Higgie, you're more and more grasping at straws.

First off, Buffett may have private data from 200 companies, but I
doubt that it's real time. Things like sales figures are at best
updated overnight, or even monthly.

Even if you assume that Buffett has real time data from 200 companies,
that's a tiny fraction of the real time data that an AI machine could
have by scanning the internet for data and patterns, and from which
valuable insights could be drawn.

You keep talking about how you're better than the bots, but you're
misrepresenting that. You're obviously better than the average bot,
but if 45% of the bots were better than you are, then you would still
beat the average because you're better than the other 55%.

These bots you're competing with are not based on sophisticated neural
network algorithms. They're written by guys in back rooms. I could
sit down with you for 2-3 hours and pick your brain and write a
"Higgie in a box" algorithm that would do what you do. But you would
still beat it in actual practice, because I could only capture a small
fraction of your sophistication in a few hours. And I'll bet that
that's the way the majority of online bots were written. That's why
they're so easy for you to beat.

The question is not whether you can beat an average bot. It's whether
you can beat the best bots, the ones implementing advanced AI
techology. Are you able to beat Tudor's "Paul-In-A-Box," the one you
referenced on June 7? I would suspect not.

There are a lot of people who dismiss the Singularity, and they always
give reasons like "A computer could never do X." That's like saying
that humans can never be better than apes because humans can't swing
from trees. And yet, humans are running the world, not apes. There's
a lesson to be learned there.

> The tools computers have are great for playing self-contained
> rules-based games like chess that have a finite number of
> possibilities. Humans suck at those kinds of tasks.

Every task has only a finite number of possibilities. The more
possibilities it has, the better computers will be at it.

> By getting computers involved in tasks that they have no business
> being involved in and for which their involvement actually confers
> massive disadvantages which are disruptive to the functioning of
> free markets, supply chains are going to collapse and the world is
> going to go into a dark age. This is as certain to me as the sun
> will rise in the east tomorrow morning.

Now you're saying something quite different. You're no longer
claiming that computers can never be better than humans at playing the
market. Now you're agreeing that they can be better, but that they'll
be so disruptive that the world will go into a dark age. Well, I
can't agree with the last part, but I do agree that computers will
certainly be disruptive.

[Higgenbotham]

The question is not whether you can beat an average bot. It's whether
you can beat the best bots, the ones implementing advanced AI
techology. Are you able to beat Tudor's "Paul-In-A-Box," the one you
referenced on June 7? I would suspect not.

You can probably figure that out based on the volume of money I was churning per day and the average rate of return per unit volume in the results that were posted. If I recall, the net return was somewhere near 0.2% per unit churned. I wouldn't be able to find a way to churn all the cash maximally as allowed by CFTC rules and still stay within risk tolerances and optimal bet size; I was probably churning about 25% of it per day on average. That turns out to be 0.05% per day and assuming 250 trading days per year comes to a compound annual rate of 13%. I think that's conservative because the actual return was more like high teens.

I'm not the best human trader in the world, or even close. Even the fact that I can beat the average bot is notable. It's even more notable that I beat the average bot playing their own game of day trading individual stocks, which I have never done. You're insinuating in the rest of this post that I'm an uninformed idiot. Yet even an uniformed idiot can beat an average bot. That's incredible when you think about it. The journalists and silicon valley really should come calling to investigate.

The original premise in starting this discussion were articles talking about what AI can do and giving trading as the primary example of what computers can now to better than humans. It doesn't really look to me like they can. I can't speak to the other examples cited, like computers can diagnose illnesses better than doctors, but I'm quite skeptical given that the primary example that was pointed to is bunk.

There are a lot of people who dismiss the Singularity, and they always
give reasons like "A computer could never do X." That's like saying
that humans can never be better than apes because humans can't swing
from trees. And yet, humans are running the world, not apes. There's
a lesson to be learned there.

I agree that would be a specious argument, as would be the argument that cockroaches are the ultimate biological survivors and were here long before humans and will be here long after humans are gone.

But my argument was different, and I've never seen it before. It was that valuing certain things like prostitutes can only be valued by humans because valuing and using a prostitute is a uniquely human experience that a computer has no use for and cannot value. You responded by saying it's well known that computers use algorithms to value equity. They could in theory use algorithms to value prostitutes too, but I don't see how that is relevant.

To carry that thought process further, if it were to be said that apes are not able to determine which trees they enjoy swinging from the most and computers are much better able to determine how much apes are enjoying the process of swinging from certain trees and how competent they feel in doing so, it would be ridiculous.

> By getting computers involved in tasks that they have no business
> being involved in and for which their involvement actually confers
> massive disadvantages which are disruptive to the functioning of
> free markets, supply chains are going to collapse and the world is
> going to go into a dark age. This is as certain to me as the sun
> will rise in the east tomorrow morning.

Now you're saying something quite different. You're no longer
claiming that computers can never be better than humans at playing the
market. Now you're agreeing that they can be better, but that they'll
be so disruptive that the world will go into a dark age. Well, I
can't agree with the last part, but I do agree that computers will
certainly be disruptive.

You would like to believe that's what I'm saying, but it's not. What I'm saying is that a situation will arise (like the flash crash only a lot worse) where it will be once again demonstrated that computers don't know how to value equity and Humpty Dumpty will not be able to be put back together.

[Higgenbotham]

"We have far too many companies in this group to comment on them individually. Moreover, both current
and potential competitors read this report. In a few of our businesses we might be disadvantaged if they knew our
numbers. So, in some of our operations that are not of a size material to an evaluation of Berkshire, we only
disclose what is required. You can find a good bit of detail about many of our operations, however, on pages 80-84.

I can’t resist, however, giving you an update on Nebraska Furniture Mart’s expansion into Texas. I’m not
covering this event because of its economic importance to Berkshire – it takes more than a new store to move the
needle on Berkshire’s $225 billion equity base. But I’ve now worked 30 years with the marvelous Blumkin family,
and I’m excited about the remarkable store – truly Texas-sized – it is building at The Colony, in the northern part of
the Dallas metropolitan area.

When the store is completed next year, NFM will have – under one roof, and on a 433-acre site – 1.8 million
square feet of retail and supporting warehouse space. View the project’s progress at http://www.nfm.com/texas. NFM
already owns the two highest-volume home furnishings stores in the country (in Omaha and Kansas City, Kansas),
each doing about $450 million annually. I predict the Texas store will blow these records away. If you live anywhere
near Dallas, come check us out.

I think back to August 30, 1983 – my birthday – when I went to see Mrs. B (Rose Blumkin), carrying a
11⁄4-page purchase proposal for NFM that I had drafted. (It’s reproduced on pages 114 - 115.) Mrs. B accepted my
offer without changing a word, and we completed the deal without the involvement of investment bankers or
lawyers (an experience that can only be described as heavenly). Though the company’s financial statements were
unaudited, I had no worries. Mrs. B simply told me what was what, and her word was good enough for me.
Mrs. B was 89 at the time and worked until 103 – definitely my kind of woman. Take a look at NFM’s
financial statements from 1946 on pages 116 - 117. Everything NFM now owns comes from (a) that $72,264 of net
worth and $50 – no zeros omitted – of cash the company then possessed, and (b) the incredible talents of Mrs. B,
her son, Louie, and his sons Ron and Irv.

The punch line to this story is that Mrs. B never spent a day in school. Moreover, she emigrated from
Russia to America knowing not a word of English. But she loved her adopted country: At Mrs. B’s request, the
family always sang God Bless America at its gatherings.

Aspiring business managers should look hard at the plain, but rare, attributes that produced Mrs. B’s
incredible success. Students from 40 universities visit me every year, and I have them start the day with a visit to
NFM. If they absorb Mrs. B’s lessons, they need none from me."

Warren Buffett, Verbatim from his 2013 Annual Report

This demonstrates some of the the intangibles Buffett relies on to value equity.

[aeden]

https://www.youtube.com/watch?v=Zz_YGeAsbBU
Good luck in July. I think swaps in hkd and usd will factor somewhat going into this window
other than the usual suspects. Take care all.

[Higgenbotham]

I've already demonstrated that I can beat the bots at their own game where they are advantaged by the exchanges by placing their computers in close proximity and front run orders, as well as place fractional bids that I am not allowed to. Those bots were not built in 3 hours in a garage either.

But now let's switch to the real game of how equity is valued. What aedens calls know the company better than your wife. You might go out to a company like Nebraska Furniture Mart and experience it as a customer, maybe like Buffett did. You experience that in the unique way humans experience things that computers never experience: the sights, sounds, buzz, feeling. You meet the employees. Maybe like Buffett, you think about the long term, maybe in your case about your retirement, another uniquely human experience. Maybe you see that the equity is selling for an average price but you meet more employees and then the owner and feel they are above average. You take a deep breath and decide to part with some of your cash and buy some equity in this company for the long term.

Good luck in July. I think swaps in hkd and usd will factor somewhat going into this window
other than the usual suspects. Take care all.

Yep, good luck and take care.