Re: [tlug] Is Japan closer to U.S. or ...?

["Stephen J. Turnbull" <stephen@example.com>]

Darren Cook writes:

 > > The point of the questions is that without more justification than a
 > > textbook ... 
 > 
 > Once you've read Ilya's book please post your list of his technical
 > errors. (The book is free,
 > http://chimera.labs.oreilly.com/books/1230000000545 )

I didn't say or imply that there were technical errors in his book,
and the right person to ask for an evaluation is Jim.  My point is
that (in another version of the Knuth quote) "97% of the time" the
bottleneck is elsewhere and effort on optimization is wasted.

Knuth goes on to say: "Yet we should not pass up our opportunities in
that critical 3%. A good programmer will not be lulled into
complacency by such reasoning, he will be wise to look carefully at
the critical code; but only after that code has been identified."  You
started by saying you weren't fussy about how "close" is measured, and
gave both bandwidth and latency as possible measuring rods.  But the
two are pretty close to uncorrelated.  So whatever your need for
proximity is, you have a 50-50 chance of choosing a measure that is
uncorrelated with satisfying your need.  IOW, you didn't identify (in
the sense of telling us, anyway) the 3% of critical performance where
optimization is valuable.

 > Over in your day job, Stephen, have you read Expected Returns by
 > Antti Ilmanen?

No.  I have nothing against Messr. Ilmanen personally, but I don't
much read about financial markets.  There's an enormous amount of
literature both technical and popular, and the sample I've seen is
crap.  I take it as an article of faith (and I mean that; it's part of
the economist's creed but is hard to prove at best) that the markets
themselves work as well as they can (except for the huge amount of
fraudulent advertising and self-serving advice generated by Japanese
securities brokers, not to forget American investment banks -- and the
worst are not necessarily bankrupt yet :-( ).  But I have some
experience with papers by financial economists, and don't think they
know anywhere near as much as they claim.  Their statistical
techniques are mathematically sophisticated but the claims of
applicability to real data are never substantiated by successful
prediction in a valid sample.  So it's all "post hoc ergo propter
hoc", and without causality there's no basis for prediction.

For example, ten years of better than market returns by a single fund
is not evidence that successful prediction is possible.  At any given
time there's only about one such company out of many thousands of such
funds, and the probably that any given fund will achieve such
performance is 1 in 10000. :-)  The only "financial economist" whose
writings I trust is Daniel Kahneman, and he's a research psychologist
and doesn't write about financial markets at all.[1]

I have little practical experience because professors in national
universities basically have taken a vow of poverty.  It's comfortable
poverty, but buying real estate in this country lives little leeway
for the necessities of life (FVO necessary == TLUG nomikais), let
alone playing in the financial markets.

 > Similar styles: a stubborn desire for thoroughness mixed with a
 > love of backing things up with actual numbers whenever they
 > can. Both authors are also way cleverer than me. :-)

Cleverness is not a virtue in the conduct of science, unfortunately.
I'd be better served with half the cleverness and twice the diligence,
not to mention a dollop of cruelty in handling students. ;-)


Footnotes: 
[1]  "Financial economist" is an inside joke.  His _Thinking, Fast and
Slow_ is highly recommended for understanding why belief in successful
prediction is so stubbornly hard to eradicate despite the complete
lack of scientific evidence that anyone has ever been able to do it.
Even J. M. Keynes was probably just lucky in picking stocks for Oxford.