it isn't my first post. i don't know what that is about.
but the GARCH comment was from one of my books, not a previous elitetrader post.
but the GARCH comment was from one of my books, not a previous elitetrader post.
Question for all of you actually making a living trading options
- Thread starter monkeyjoe
- Start date
if this is your post 1 how could you have written a previous:i just want to clarify what I mean by "every options trader basically thinks of the world as a GARCH model", which i doubt is what i actually wrote.
Martinghoul - can you elaborate? You mean scenario analysis like volatility cones? ...or just many repeated simulations with tons of different assumptions.
sonoma - I have a PhD in economics and am a professor (non-finance) by day. I don't believe in most of the models in many ways, but I do see them as an occasionally useful tool (see my comment on mean reversion in the first post as an example of this). My main interest is in plain vanilla equity options... these are what I have access to as a retail trader, and they are also liquid enough/watched enough that meaningful data are available.
filthy (Mr. S - great books! I've seen your posts on NP) - that was a dramatic oversimplification of what you wrote. I am pretty sure you wrote that GARCH can be useful in some settings but has a ton of limitations and is not really believable for modeling fluctuations in Vol in the marketplace. So... sorry for that - I was mostly just trying to spur some modeling discussion.
That's easy - at some point risk management kicks in and you don't want to have anything to do with it. Truth is, everyone who "helped" to cover their positions made out like a bandit - I was trading IR swaptions at the time and it was true feast.
It's just an ample description of my life, not real quote value.
That's where I started. Lately, I found that it's more "interesting" to view the opportunities in volatility space as in terms of relative risk premia. It is easier to understand whose lunch you are eating this way, which, in my view, is very important.
It is not clear to me that short term traders are creating any real liquidity or value, just like the high frequency fund creates a lot of volume but little real liquidity.
A investor that buys companies based on their fundamental worth is adding some value by facilitating capital markets, but a successful investor is compensated far more than the value he adds to society.
If anything the traders that like to buy stocks that are "working" seem to perpetuate market bubbles like TSLA or Bitcoin and destroy value for society.
yes...you probably gather mine was a rhetorical question to him....because if that was their only problem it was easy solved!
Sounds like you had fun.
It was times like that whereby my view on a lot of these models came from. That at the end of the day it is when prices are well outside the model, you know the reason and then you can provide liquidity. Thats opportunity!
Similar to posts you have made regards selling OTM puts - when there is value in doing it it makes sense as there is no natural supply for whatever reason and prices are high.
Alternatively sometimes having to pay a price when you can for liquidity - as you finally get a seller - even if a model is saying its expensive can also be valuable.
Do you know much of the London Whale - all I know is what I read and it seems there was a little bit of this occurring there?
Definitely the latter...
I am stuck at home sick, can't even walk my dogs so this is as close to entertainment as I'd get today 
This one is interesting. Distribution bets (terminal only) are very different in nature vs actual volatility bets where you do hedge delta. Given that options are more or less priced in the risk-neutral world, you can sometimes find a terminal distribution trading strategy that would give you a statistical edge. There is a variety of examples, such as roll-down trades in futures.
Isn't this also kind of the same point you are making in recent posts about regulatory arb and selling naked index crash risk?
I think it's really interesting to see the ways in which people explain the overpricing of crash risk. There are those who say it's regulatory arb, those who say it's a natural demand-for-insurance/risk aversion which drives the prices up, and there are others who say it reflects the vol spike in the risk neutral distribution as prices fall rapidly (which is an argument that could hold even in the unlikely case you are able to hedge all the way down). I'm guessing the truth is probably a combination of all three. Still, it seems to me that a lot of clever replication arguments have to devolve into terminal distribution arguments when atypical events are being addressed.
Well, not really. Short-dated tails are usually overpriced both from the terminal distribution and from the vol perspective. The terminal distribution trades can be something where you thing real world drift is statistically mispriced - for example, I've been investigating if forwards on high-dividend ETFs are mispriced because the dividends accrue like they do for bonds and the ETF should actually increase in price (dirty price sort to say).
My approach to modelling is actually very tactical - I usually assemble a few simple models (either relationship-based or actuarial) for a particular strategy and try to combine the results into a single indicator. This strongly contradicts the holy grail modelling approach, but I found that tractical relative value models plus scenario tweaking using simple black scholes (but with realisitic scenarios) is sufficient to produce very reasonable results.
Also, unless you are selling tails, you can never have extremely good sharpe trading options (I'd say Sharpe of 2 for a portfolio of 3-5 strategies is stellar) but if you do things right you can also have very persistent stream of p&l since risk-premium related edges last pretty long time.