Author Topic: Ed Thorp on Statistical Arbitrage, from Wilmott.  (Read 42007 times)

Offline kido

  • SBC Old Boys
  • ***
  • Posts: 310
Re: Ed Thorp on Statistical Arbitrage, from Wilmott.
« Reply #10 on: 12 September 2009, 10:51:50 »

Google or Wiki the term.

When a betting system ask you to increase the bet amount solely on the condition of lossing the last bet, and/or decrease to a smaller amount after a winning bet, this system is generally said to be "Martingale". And most of the Martingale systems are applied to independent event, which make it more ridiculous.

Call me to arrange a time to pick up the books.

Wiki-ed. I see. This strategy was claimed by many rookie gambler that they could beat the casino.  I remembered I was told by an uncle of mine (when I was really small, in primary school) that this method could beat any casino. ( In fact his strategy was to play 大細 and to double his bet everytime he records a loss, until a win occur, and to stop playing after this win.).

At that time I thought he was 'quite right', the prob. of getting a 'win' in long time will be 100%. So in long time he must be win.  But in reality this is impossible.  He will run out of cash shortly.


Hey, diddle, diddle ! The cat and the fiddle.

Offline kido

  • SBC Old Boys
  • ***
  • Posts: 310
Re: Ed Thorp on Statistical Arbitrage, from Wilmott.
« Reply #11 on: 12 September 2009, 11:17:08 »
Google or Wiki the term.

When a betting system ask you to increase the bet amount solely on the condition of lossing the last bet, and/or decrease to a smaller amount after a winning bet, this system is generally said to be "Martingale". And most of the Martingale systems are applied to independent event, which make it more ridiculous.

Call me to arrange a time to pick up the books.

Sidetrack a little bit: There is a strategy in game theory that your strategy depends on the past result, consider the prisoner dilemma (I always mis-spell this word), a good (or best) strategy is a very simple strategy called tit-for-tat, which decides the next move by the result of previous moves. 

Both Martingale/Kelly decides the amt to bet on next move/bet.  Kelly based on the estimated prob., Martingale based on pass results.  That triggered me to think that maybe Kelly could/should be applied on those games that are perfectly probabilistic, like those in casino.   And Martingale should be applied on games where 'human factor' is important, like the stock market.
Hey, diddle, diddle ! The cat and the fiddle.

Offline chin

  • Global Moderator
  • *****
  • Posts: 6,663
Re: Ed Thorp on Statistical Arbitrage, from Wilmott.
« Reply #12 on: 12 September 2009, 11:50:26 »
Sidetrack a little bit: There is a strategy in game theory that your strategy depends on the past result, consider the prisoner dilemma (I always mis-spell this word), a good (or best) strategy is a very simple strategy called tit-for-tat, which decides the next move by the result of previous moves. 

Both Martingale/Kelly decides the amt to bet on next move/bet.  Kelly based on the estimated prob., Martingale based on pass results.  That triggered me to think that maybe Kelly could/should be applied on those games that are perfectly probabilistic, like those in casino.   And Martingale should be applied on games where 'human factor' is important, like the stock market.

IMHO, Martingale shouldn't be used at all. Look at my comment on the LTCM.

Kelly is good enough to be used in the investment markets, at least that's what Ed Thorp says in the articles. Especially in quant based trading where you can establish probabilities.

Offline chin

  • Global Moderator
  • *****
  • Posts: 6,663
Re: Ed Thorp on Statistical Arbitrage, from Wilmott.
« Reply #13 on: 17 September 2009, 03:14:52 »
I just did a Wordle of the main text in this thread. (www.wordle.net)

Can one tell the main idea from just looking at the picture??!!

Offline kido

  • SBC Old Boys
  • ***
  • Posts: 310
Re: Ed Thorp on Statistical Arbitrage, from Wilmott.
« Reply #14 on: 17 September 2009, 12:45:28 »
I just did a Wordle of the main text in this thread. (www.wordle.net)

Can one tell the main idea from just looking at the picture??!!
This is kind of a semantic map, periodically I think the way Google arrange/manage their information is like this. (Although I have no prove). And in A.I. this kind of diagram is quite oftenly used.

Back to the diagram: imagine of the size of the word is the freq. it occurs, the distance between the words are the semantic/meaning they're refering,(Here I am imagining wordle got an A.I back end to 'understand' the meaning of these words, but it seems not, they're closer because they're simply closer in the passage context, e.g. 'Kelly' & 'Criterion') hence the closer the words are, the closer their meaning.

As a result : "betting" and/or "statistical" is/are the centre of this passage, everything is going around these 2.
Hey, diddle, diddle ! The cat and the fiddle.