Pair Trading & Low Frequency Trading

aks1989 · Post by **aks1989** » Tue Oct 13, 2015 9:04 pm

Hello Guys,

I want to ask one question, does the pair trading works in low frequency trading.

In high frequency and ultra high frequency, I can think of a reason behind pair trading. Like, there can be deviation for one symbol from normal price movement because of some improper pricing in very short term or because of high demand and supply, price of one symbol deviated and many more reasons.

But in long term, is pair trading profit possible. I don't think of a logical reason. Because if there is a deviation, then this can be removed by high frequency trading. So, is there any reason of pair trading in low frequency trading?

MCgeneratedname · Post by **MCgeneratedname** » Tue Oct 13, 2015 10:19 pm

> if there is a deviation, then this can be removed by high frequency trading

You assume that the only direction for a deviation is to close over time.

How do you know that the deviation won't extent?

aks1989 · Post by **aks1989** » Tue Oct 13, 2015 11:05 pm

Yes, that's what I want to argue. Why deviation will extent over time.

Note : here deviation means that one of the cointegrated pairs deviated from the cointegrated series while other is still following the same trend.

I think that if there is deviation, and assuming market knows about the pair, then high frequency will start making trading around the deviation till the two time series again merged to follow cointegrated.

What do you think?

MCgeneratedname · Post by **MCgeneratedname** » Tue Oct 13, 2015 11:32 pm

I have given up answering "why"-questions. They started driving me crazy, or should I say crazier ${L_SMILIES_WINK}$ .

I think it is more fruitful to acknowledge that there are forces that pull pairs apart and back together.

There is always the possibility of a structural break in cointegrated time series, i.e. they are not cointegrated anymore. If you are able to detect the break then you can exploit all the mean reverse strategies that are not able to detect the break.

rickyvic · Post by **rickyvic** » Tue Oct 27, 2015 12:05 pm

Normally these forces are due to liquidity, news flows (stock of information) that affect risk premia, or simply general preference for two related assets, say two companies in the same sector, two bonds of the same issuer or two share classes of the same company, etc...

That is where arbitrageurs and leveraged investors come into place, by providing liquidity to the markets, eating up that inefficiency.

Now this can be done at any frequency, obviously the more the expected time to restore the equilibrium the more the inefficiency has to be "not easily detectable".

Also to take long term positions you need to take more risk, so more capital to commit to the trade and many short term traders do not have enough or are not willing to.

AB12358 · Post by **AB12358** » Tue Oct 27, 2015 12:11 pm

Does anyone know of any studies done attempting to quantify just how much more 'efficient' this has become over time? (ie. impact of stat arb and pairs trading)

umuzungu · Post by **umuzungu** » Mon Nov 09, 2015 12:55 am

Here are some papers that try to measure the effectiveness of pairs trading in US equities. If anyone knows of more recent work, I'd love to hear about it...

Avellaneda & Lee (2010). Statistical arbitrage in the US equities market. Quantitative Finance 10(7), 761-782.

Do & Faff (2010). Does simple pairs trading still work? Financial Analysts Journal 66(4), 83-95.

Clegg (2014). On the persistence of cointegration in pairs trading. http://ssrn.com/abstract=2491201