Yay my 100.00000000000000000000000th post!
im just looking around if this is viable. im reading some stat arb in stock market. basically finding highly correlated pairs, according to arbitrage pricing theory similar ones should have same prices in the market, and any deviations from another will collapse to the single same price. so we exploit this.
but note, stat arb is not same is arb where theres riskless profit. stat arb simply revolves around the idea that prices of similar securities/instrument should be same, or revert to it if different at some point. there are many risks involved, however its attractivness lies in the market neutral nature of the strategy.
however, i know that currency pairs in FX market can vary time to time depending on the time frame. like the jpy pairs, they appear to be strongly correlated from one another at least in shorter time frames (theres a whole list of table of correlation in Kathy Lien's book can't remember em all but shows that correlation changes drastically from - to + in various time periods).
also, stat arb in currency pairs would be a bit more of a releif as you dont have 5000 stocks and 17 million possible pairings to test on.
anyways, let me know what you guys think, i would like to find out more.
im just looking around if this is viable. im reading some stat arb in stock market. basically finding highly correlated pairs, according to arbitrage pricing theory similar ones should have same prices in the market, and any deviations from another will collapse to the single same price. so we exploit this.
but note, stat arb is not same is arb where theres riskless profit. stat arb simply revolves around the idea that prices of similar securities/instrument should be same, or revert to it if different at some point. there are many risks involved, however its attractivness lies in the market neutral nature of the strategy.
however, i know that currency pairs in FX market can vary time to time depending on the time frame. like the jpy pairs, they appear to be strongly correlated from one another at least in shorter time frames (theres a whole list of table of correlation in Kathy Lien's book can't remember em all but shows that correlation changes drastically from - to + in various time periods).
also, stat arb in currency pairs would be a bit more of a releif as you dont have 5000 stocks and 17 million possible pairings to test on.
anyways, let me know what you guys think, i would like to find out more.
{Promotion Removed}