The usage of Delta Neutral strategies, by its very nature, would tend to imply that one is not particularly keen to have a view on price, otherwise they would go for the directional trades. Volatility on the hand, because of its mean reverting tendencies, can, imo, be forecast (for one of a better word) more reliably then prices could be. So times like now when IV is fairly low (for the indices), spreads could be put on which are +vega with a slight -delta bias. That way when the big moves in the underlying take place, the +vega can be counted upon to provide some assistance to these +theta strategies. Workable strategy ??
Delta neutral trading
- Thread starter nerdzkilla
- Start date
The usage of Delta Neutral strategies, by its very nature, would tend to imply that one is not particularly keen to have a view on price, otherwise they would go for the directional trades. Volatility on the hand, because of its mean reverting tendencies, can, imo, be forecast (for one of a better word) more reliably then prices could be. So times like now when IV is fairly low (for the indices), spreads could be put on which are +vega with a slight -delta bias. That way when the big moves in the underlying take place, the +vega can be counted upon to provide some assistance to these +theta strategies. Workable strategy ??
- Joined
- 22 July 2009
- Posts
- 602
- Reactions
- 0
I am assuming we are referring to a delta neutral position now rather than a
delta and vega neutral portfolio (as I was assuming before)?
Generally yes, depends on what your 'slight delta bias' is. E.g. Trading a calendar where it is minimally otm will likely have gamma/dgamma losses exceed +vega gains, despite your view of vol being correct.
How much faith should be put in mean reversion of vol? Just to stir the pot, if a basic GARCH(1,1) is fitted to analyse index vol, suggesting clustering effects dominate mean reversion tendencies.
The one million dollar question...
To me it seems mean reversion is quite good from high vol to the mean, but much less reliably so from low vol to the mean.
- Joined
- 18 June 2010
- Posts
- 277
- Reactions
- 0
ah, you have just summarised in one short sentence what I just spent 2 paragraphs typing out.....
- Joined
- 22 July 2009
- Posts
- 602
- Reactions
- 0
- Joined
- 18 June 2010
- Posts
- 277
- Reactions
- 0
first thought is , well VXX = VIX, since VXX is a combination of the front two months VIX futures, so you would have to find some sort of pricing anomoly to make that profitable.
second thought; assuming he is talking about same month (since he talks of both expiring together), it gets more interesting. A bet on Dec VIX is a bet on where spot VIX ends up on the xth december. A bet on Dec VXX is a bet on where VXX ends up on xth December, but because of the way VXX rolls from the front month to the next available month daily, on xth dec expiration day VXX would consist 100% or near to it of Jan VIX futures (or maybe Jan + Feb)
net result
long VIX calls => long Dec vix
short VXX calls => short mainly Jan vix
When I can find a decent model or some serious size? My options model is still the same, just long vanilla straddles when it thinks the time is right otherwise in cash.
I have one good idea for vol arb (slightly unconventional I believe) but it needs big stakes behind it. I found a good collar model but the returns are quite similar to my regular model.
- Joined
- 18 June 2010
- Posts
- 277
- Reactions
- 0
when i can find anyone rich and dumb enough to give me 20% of winnings without penalty for losing, for executing a pretty basic strategy. Then what would I care whether its profitable or not........
wayneL
VIVA LA LIBERTAD, CARAJO!
- Joined
- 9 July 2004
- Posts
- 25,578
- Reactions
- 12,705
- Joined
- 22 July 2009
- Posts
- 602
- Reactions
- 0
Sorry, not many.
I haven't modelled the term structure of VIX futs extensively, but there is definitely no arb between VIX/VXX calcs.
Considering Dec and Jan tend to be quiet/stagnant, the term structure will generally hold as upward sloping, so would fulfil his second critical factor listed. I had a quick glance at recent VIX futs, Nov and Jan were priced higher than Dec.
Seems a solid idea, but would have to check the persistence of this cycle and effect on the calls due to differing expiry dates of the ops.
DeepState
Multi-Strategy, Quant and Fundamental
- Joined
- 30 March 2014
- Posts
- 1,615
- Reactions
- 81
Any guru out there trading gamma on equity indices or FFX?
Hi,
When they talk about delta neutral strategies, they are talking about removing the risk of market up/down moves to your portfolio. To do this, you will have a mix of bullish/bearish plays in your portfolio e.g Sell Call on Apple (Bearish), but Sell Puts on the NASDAQ (Bullish) - note, maintaining delta neutral is an ongoing thing, as this will change as your underlyings move.
However, your main source of profit in options let's say, would be from theta decay (time decay) or collapse of vega (Volatility). However, in this strategy, you need to ensure that you practice good bankroll management, and also watch your vega risk. To do this you will need a mix of strategies.
Returns will vary depending upon your tolerance for risk, but consistent returns between 15% - 35% would be realistic.
Hope that helps.
DeepState
Multi-Strategy, Quant and Fundamental
- Joined
- 30 March 2014
- Posts
- 1,615
- Reactions
- 81
Are you active in this space or are you delta or vega oriented? I am interested in harvesting the premium for neg skew available from neg gamma with delta hedged via algo. For me, neg skew with controlled risk exposure in a diversified arrangement is easy to bear..taking into account conditional correl effects. All positions would be massively overcollateralised. Overall, I see this as extracting another risk premium apart from equity risk and other smart betas. Would be good to have a thought partner if interested.
Hi,
I trade predominantly short call/put options in u.s market around the 1.2-1.3 S.D OTM points and 55-45 DTE. I place greater emphasis on probability of ITM/OTM and volatility than I do with other greeks. I keep a mental note of deltas (negative at the moment, not trying to stay neutral). By default, I'd have a low gamma risk because of my entry & exit points (out of trades usually >15 DTE), but don't use it as a measure/filter. Volatility skew does not concern me either, I'd still sell the other side as it does not require additional margin and P&L appears to be positive despite POP decreasing.
I rely on third party sites for information as I don't have access to options data to back-test and don't know what programs to use.
DeepState
Multi-Strategy, Quant and Fundamental
- Joined
- 30 March 2014
- Posts
- 1,615
- Reactions
- 81
Thanks. To repeat my understanding:
So you are neg gamma if not intensively so. Write options with presumably variable ratios on call and put wings OTM at inception and roll forward via calendars to maintain approximate maturity. Target deltas at inception (ie look at ITM/OTM) but are not tightly managing these as a book subsequently. Happy to sell into skew....noting that this is profitable (this is where we are highly aligned, if the actual implementation is not identical).
Bottom line appears you are seeking to profit from theta decay and opening positions based on a set of rules related to initiation delta. Although you keep an eye on vol, I am not sure what you are doing with this information. I do not know how you find mis-valuation as opposed to using the above to shape the payoff.
If this is correct, a key difference between what I am seeking to do and what you are doing is tightness of delta hedging. For discussion purposes right now, the actual positions mean less to me than the greek outcomes which are broadly similar apart from the fact that you are happy to let delta roam. If gamma got too large and pin risk became an issue, I would roll. In reality, I look for discontinuities in the vol surface that were not well explained for position initiations and try to get a desired strategy shape across a range of metrics.
I built my own backtest engine back in the day. MatLab infrastructure. I presently have a blank sheet of code space waiting to be filled again if I go ahead with this. The underlying method for this purpose will be Black's approximation for the most part, but I will also build in a binomial lattice in case I need more flexibility for GARCH effects or events.
For my purposes, if IV exceeded realised vol, a profit was made as everything would be hedged continuously etc. Obviously this is an approximation, but one which is adequate for my purposes. The differences between the approximation and reality are what creates the edge. There is almost always be a bias in IV and skew due to risk management concerns and frictions. I am seeking to arb that because I do not face specific elements of those concerns in the same way and am more prepared to accept basis given smaller relative position sizes, available cross-sectional diversity and propensity to accept time diversification. The underlying argument is hardly rocket science and is simply one relating to the price for risk bearing.
I have history and live data on active series. I do not have it for historical, inactive, series. Backtests would need to be of the type that uses current options pricing and historical market developments. Kind of like shock tests based on historical data. It would also give an idea of the distribution of return outcomes for a given strategy, even if the mean is off. Backtests for options are quite different in purpose than for other types of equity strategies for this strategy type so full history is not actually needed by me if I have VIX vs realised. The rest is risk management.
If you wish to collaborate at some level, the key areas of overlap in terms of objectives seems to relate to viewpoints on what represents value in the outer wings to neg gamma traders. Position sizes may also vary depending on views on what value is inherent in the current level of IV, skew and calendar effects. Possibly, you may have a wish to share/exchange views on re-hedging delta. I will do this via algo, but there will be acceptable basis risk which can be tightened with discretion. S&P 500 would be a good enough place to start as any.
It would be great to get an active forum of less directionally oriented option traders going here. Please let me know if some form of collaboration might be useful to you (and others...)
I'm more likely to sell premium in trades with a higher implied volatility percentile/rank vs. the year's range and will have more % of total capital invested when VIX > 15 than say <15 (~60% vs 30%) - these allow me to go further OTM. If I feel I have too many short vega on in low IV environments, then I will think twice about selling premium on downside or I'll either not put on trades OR buy a spread/calendar/diagonal.
Average monthly Implied volatility in general has exceeded historical volatility over the last 10 years (except 2008) - it's the 2008 incidents that I'm always conscious of. I don't have any hard fast rules, I choose whatever strategy that suits me then. The only hard fast rule that I follow is not letting my options expire, and taking my winning trades off about 30-50% range. I keep things simple, so I can sleep at night.
Similar threads
