Search results

Come to think of it, there are two distinct flaws, both bad but a combination is truly deadly. Flaw 1. “I am the smartest guy in the room” Flaw 2. “I am right and will fight instead of listening”

I am not using IB, for obvious reasons. For brokers that do some pass-through, in most cases the “bid-ask” on these rates (borrow vs rebate) is like half of the carry. I am sure IB is not an exception.

Some people do it systematically just like that. Their reasoning is that the risk premium is there all the time, so you keep harvesting it. In the world of limited capital, it’s less about mispricing and more about risk reward - for example, SPX vol in the current environment (pricing 7 and...

I doubt a relational db would be an improvement speed-wise - pretty sure my data frame fits into memory and pandas is pretty efficient. I just don’t know how do implement it properly for specific types of searches

Yeah, fees do add up. However, anything OTC is a PIA in many aspects - eg on an unwind you might have dealers trying to take advantage and have to deal with novation

LOL

Well, while that's true in theory, it's not exactly true in practice. Funding and borrow costs differ across the platforms, while if there is an exchange-traded product it's far more transparent. Also, futures provide long/short symmetry with regards to borrow and funding costs - right now, if...

People were planning to quote variance futures too and you know how that turned out :)

Personally, I think it will fail. For a number of reasons, main one being cash settlement.

It's a fairly common strategy. Simplest form you can go long dividend select spiders and short the futures against it. It's something that you can make arbitrarily complex - you can go down to sectors or what not. There is a number of interesting things to do around dividends.

In 2005-2006 variance swaps were the shit, you can trade them super-liquidly. I recall getting 5-7bp wide markets. There were funds who did almost only variance swaps. Crisis put an end to it, with counterparty risks and all. Nowdays, a buyside client can get it about 7-10bps wide, which is...

Pandas has wonderful features for that - in fact, I can even do something like cond = (opt['exp']=='20170907') & (opt['cp']=='C') x = opt[cond][abs(opt['delta'] - targetDelta) == abs(opt['delta'] - targetDelta).min()] However, my main issue is that when I am searching through a lot of options...

If I want it to work fast, would I want to add these fields to the index? I tried but it behaves quirky and, oddly enough, we don't have a python expert in house (at least none that I can talk to).

Winnie The Poo, of course.

Lets say I have a data frame with options data. I would like to be able to quickly find an option on a given date with a given type, expiration that has delta closest to x. What’s the best way to do that?

I am not a fund manager, just a lowly vol arb PM. However, the investor in this fund (its a single investor) as well as investors in the funds I’ve worked for before are large institutional investors who knew the business and understood the risks. Could you say that about your clients...

He does not have to. Anything that’s not a core expertise of the fund is outsourced. He does not even do his own stock accumulation. Most of the things you mentioned are occasional and you don’t have to have a person in-house for it - an investment bank can help with it just as well. I think we...

That’s not his business model. He’s an activist fund which means he takes large stakes in companies and creates positive value by changing those companies (management, capital structure etc). That’s what he’s arguably good at and it would be silly for him to change that. Predictably, it produces...

Are they going to list options on this thing?

Looks like there is real size trading there :D