CORRELATION EminiS&P-Russel2000

[abogdan]

Quote from mind:

abogdan, riskarb

dont waste time. its just a word. sure it is not true arb. some kind of stat arb it is for sure ... in the end the question is if it works.

abogdan, what is your sharpe ratio expectation on this system? above 1.5? i guess so given the mentioned trading frequency.

peace

Yes, that is about right. Mind you though (no pawn intended) that we only make 0.5 point on average a day per two contracts. Not a lot, but we keep this system alive for one of our small accounts.
It also requires constant readjustments of the algorithm which is also somewhat painful. We have much better pairs to trade than this. We also run Merger Arb system that generates just around 1% a month with virtually no risk.

 

[gekko]

Quote from abogdan:

Yes, that is about right. Mind you though (no pawn intended) that we only make 0.5 point on average a day per two contracts. Not a lot, but we keep this system alive for one of our small accounts.
It also requires constant readjustments of the algorithm which is also somewhat painful. We have much better pairs to trade than this. We also run Merger Arb system that generates just around 1% a month with virtually no risk.

Hello abgan,

I have been making a few trades this past week on the ES/ER spread. For instance, yesterday at about 11:40, I did the following transaction:

Spread difference: 530.50
Long ES @ 1099.00
Short ER @ 568.50


Then at 13:00, when the spread increased by three points I did the following transaction:

Spread difference: 533.50
Close ES @ 1103.50
Close ER @ 570.00

Total profit per spread: $75 not including brokerage fees.

Is it the right way do do it. I was expecting to make $50 per point, instead of $25.00

Am I doing it the right way ?

Thanks

 

[abogdan]

Quote from gekko:

Hello abgan,

I have been making a few trades this past week on the ES/ER ...

We typically get in when the spread on the adjusted values is bigger than 2 standard deviations calculated on 30 days sliding window with the mean calculated on 15 min bar chart. (Simple Lin Reg Value). We close the trade when the spread collapses to about "0".

 

[stock777]

Quote from mind:

abogdan, riskarb

dont waste time. its just a word. sure it is not true arb. some kind of stat arb it is for sure ... in the end the question is if it works.

The road to ruin is littered with those that thought it was just a word.

 

[mind]

Quote from abogdan:

Yes, that is about right. Mind you though (no pawn intended) that we only make 0.5 point on average a day per two contracts. Not a lot, but we keep this system alive for one of our small accounts.
It also requires constant readjustments of the algorithm which is also somewhat painful. We have much better pairs to trade than this. We also run Merger Arb system that generates just around 1% a month with virtually no risk.

i c. i like that way of thinking. i was not aware that currently anyone could make money in merger arb. i thought that game was over.

 

[mind]

Quote from stock777:

The road to ruin is littered with those that thought it was just a word.

now t h a t is a really short and meaningful statement. very much agreed.

peace

 

[TheStudent]

Quote from mind:

i c. i like that way of thinking. i was not aware that currently anyone could make money in merger arb. i thought that game was over.

I also thought merger arb was commoditized, but maybe someone with more experience can comment. The trouble is that merger arb people will never tell you that their business sucks.

But actually I don't think that merger arb will ever go away because mergers will never approach 100% certainty.

the difference between merger arb today and 20 years ago could be that in the past, arb players took p&l from the mispricing of closure risk by general shareholders. Now, the mispricing is more or less gone, and the p&l comes from weaker arb players making mistakes or sales of exposure to tail events.

In an efficient merger market thus, arb spreads should wax and wane with interest rates.

Just my hypothesis - anyone care to verify?

 

[stock777]

Like a lot of street games Merger arb, such as it is, will remain a tactic for eternity. Why?

Cause they are using OPM (other peoples money) to roll the dice.

 

[mind]

Quote from TheStudent:

I also thought merger arb was commoditized, but maybe someone with more experience can comment. The trouble is that merger arb people will never tell you that their business sucks.

But actually I don't think that merger arb will ever go away because mergers will never approach 100% certainty.

the difference between merger arb today and 20 years ago could be that in the past, arb players took p&l from the mispricing of closure risk by general shareholders. Now, the mispricing is more or less gone, and the p&l comes from weaker arb players making mistakes or sales of exposure to tail events.

In an efficient merger market thus, arb spreads should wax and wane with interest rates.

Just my hypothesis - anyone care to verify?

well, somehow that is what i heard. people are now playing the merger deal in and out several times before final fusion takes place.
i do not know what to think about it. requires more leverage now, i guess to make it meaningful. maybe it is just seen as a pairstrade where the merger announcement plays the "glue", everything else is technical. maybe kind of 2std bollinger - abogdan referred to something of that kind.

peace

 

[TheStudent]

my point exactly - when spreads narrow because of many players, leverage becomes more important - hence profitablity becomes more sensitive to interest rates

also, it is a dangerous short vol situation - periodically 3 sigmas are going to come along and wipe out the less diversified and heavily leveraged players. so the game is not arbs taking money from the public but arbs fighting other arbs