[ExI] Efficiency of algorithmic trading

[Eugen Leitl]

On Fri, Apr 29, 2011 at 06:10:38AM -0400, Rafal Smigrodzki wrote:

> ### Well, why should the clueless expect to gain? Inefficient actions

There's a considerable incitement to bet large fractions of your
retirement fund into stocks, at least in USia (Roth IRA etc.).
Inversely, in Europe we have a pension model which is equally a Ponzi
scheme, based on particular expectations in economic growth,
wealth distribution and demography. Both models are broken.
Of course subscribers to both models are technically suckers,
but there's at least some criminal energy present on the part
of peddlers of both models. And it's not so easy to opt out
of these, deliberately designed to be so.

> (i.e. the short-sighted greed of the small investor) lead to incorrect
> pricing (i.e. prices that do not accurately reflect the underlying
> values of assets) and better investors (i.e. investors who understand
> the future over larger time scales) correct the inefficiencies,
> reaping a reward. There is nothing wrong with it, and it's the reason
> why I do not individually trade on the stock market.
> 
> It bears repeating that the social purpose of the stock market is not
> the money that investors make there but rather the pricing signals
> generated by it, and the accompanying flows of resources to productive
> uses. Accurate prices are, well, priceless for the functioning of an
> economy, and the gains that investors make on the stock market are a
> small price the rest of us pay for the knowledge that is encoded in a
> price. Any feature of a trading system that produces more accurate
> prices is socially beneficial, no matter how many little greedy people
> get burned by it.
> 
> ### Trade *is* human (even if done with algorithmic intermediaries). I

While the prime causes currently reside in primates it is easy to
see how systems become increasingly autonomous, pursing trades in
their own interest. This can be as simple as household appliances
realtime-bidding on slices of the power pie, with bid limits
depending on how critical the function is. 

> am all with you in the decentralization camp but I haven't heard any
> interesting specifics from you yet. The fraud is mostly related to

It's basically just Tim May's vision, slightly souped up for
today's end user cloud (freedom-box-like systems). The idea is
to move critical parts of finance (like the monetary system, mint,
individual transactions, and higher level functions based on above
primitives. Reputation tracking, notarization, uncensorable permanent
public record, and such. 

> government involvement in banking, otherwise it tends to be a

You need to shield the system from manipulation on part of end
users or user groups. Of course cryptography is no panacea (and
would require diversity to avoid becoming a pivotal point of failure)
but then it's not hard to improve on the track record of federal
and corporate crooks.

> self-sanitizing business (just like other branches of the economy
> where reputation is key, and performance is easily measurable by
> outsiders).

One of the key functions if providing an unfalsifyable, publically
inspectable track record of interactions, so that nym reputation 
for the particular scope can be obtained in realtime.

Yes, most of the above are extremely hard problems. The smart
people who still think Wall Street quant is a glorious occupation
should get on that track instead.

-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
______________________________________________________________
ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE