Capitalizing on Short Term Equity Price Fluctuation Predictions
A lot of day trading activity goes into trying to predict short term fluctuations. However, wouldn't the Bid-Ask spread prevent making any profits on these short term fluctuations?(the spread on equities is often like over ten percent of the functions you can expect in a multi minute window.) I am a nub so bear with me.
IE: if you were hypothetically able to predict with like 55 percent certainty that a stock would be higher in three minutes how would you be able to capitalize on that without getting destroyed by the spread? Maybe limit orders or something?
Thanks
You said it. Limit orders. There are studies showing that ideal strategies for working an order (splitting between limit and market) while minimizing slippage do exist.
Also, it's impossible to predict t + 3 minutes or t + 3 years, but you definitely can predict t + epsilon with epsilon in (0,~ 300ms]. How is the million dollar question asked by PTFs
Thanks for the response, but how can these trades possibly be profitable? Even if you get in at a limit order then you will eventually have to pay the spread when you get out or take a hit in your profit. If you are only in for under a sec how can the stock swing enough to makeup that spread?
You have lot of different kinds of statistical arbitrage. There are maybe some folks who actually try to predict with minute bars t+3 minute, CTAs and PTFs in chi-town and they might be successful, but the classic stat arb by hft works in a different way and sadly no one is going to tell you how to do it exactly. You don't have just the best bid & ask, you also have an orderbook which entails a lot of information and fluctuates heavily. Place a limit buy order at the bid, get filled, market drops. Hit the ask. Net profit zero, but you get kickbacks from the exchange for providing liquidity. Do that 100 times a second and you gucci. You know that people actually try to detect iceberg order, so you can also confuse others by "faking" an iceberg and unloading your position on iceberg front-running algos. You can detect orderflow from retailers (who shouldn't have an edge), because they don't cancel 90% of their orders within 250ms. If a "stupid" retailer wants to trade with me, I am taking the opposite sign ofc. I will win in 55% of all cases and use my built up inventory as a prop bet. If BlackRock wants to trade with me, I am going to stay the fuck away because there might be another 10.000. contracts hitting the screen in the next 5 minutes
Investing in Volatility (Originally Posted: 01/11/2011)
I'm looking for any articles or guides that explain in a practical manner how to invest in a portfolio of options such that you obtain only a long exposure to volatility index. I've looked into the VIX Futures, but for several reasons these are not an ideal way to obtain exposure to volatility over the long-term. I've read several academic articles by Derman, Whaley, etc., and my understanding of how this works is that you must take a long position in S&P500 index options, and then delta hedge this position.
What I'm after can be boiled down to three things: 1) The specifics as to the composition of the long index option positions (what type of spreads/straddles, weighting by strike price, maturity, etc.); 2) The required frequency of adjusting the delta hedge; and 3) Overall performance of this investment as compared to the VIX.
Thanks for any help you can offer.
http://etfdb.com/etfdb-category/volatility/
Personally I like the XXV. 2011 is going to be pretty volatile, so it might be nice to own some of this.
Thanks for making me think about it.
I've thought about just using VIX futures, but there are two big issues with this approach: 1) If you use short-term futures (0-2 months), you get a decent correlation to VIX (but still not 1:1) and you almost always wind up buying at a 10%+ premium to current VIX and settling at actual VIX, so you generate a consistent loss each period.
2) If you use medium term futures (3-7 months), you don't achieve anything close to a 1:1 correlation with movement in the VIX.
These will be an issue for the ETFs which invest in VIX (which is how I believe most of them gain their exposure). That's why I'm after details on how a fund can use an option portfolio to obtain a cleaner exposure to VIX.
Sorry, why is 1) necessarily true?
I pulled data from 2004-2010 (the first year VIX futures were available), and there is a very consistent trend of VIX futures trading at a substantial premium. Although you would expect them to price based on mean reversion, and thus show no premium or discount in the long-run, the data do not bear this out. I suspect this is because there is an imbalance in demand between those who are looking for volatility to decline and those using VIX futures for portfolio insurance -- with much more volume on the portfolio insurance side.
why not invest in variance? you can easily replicate variance swaps with a portfolio of OTM options. thats static replication - no rebalancing, etc
http://www.ivolatility.com/doc/VarianceSwaps.pdf
[quote=JimSimons]http://www.ivolatility.com/doc/VarianceSwaps.pdf[/quote]
My understanding of variance swaps is that you do need to dynamically hedge such a portfolio of options as the underlying index moves (i.e., when the index moves up, you close out some calls and buuy puts instead)?
you would not need to delta hedge the variance swap replicating options. However, you need to ideally buy a portfolio of options across all available strikes weighted by the inverse of the squared strike. Sometimes, this is not very practical/very costly due to illiquidity or just the shear volume of options for a non institutional investor
1) Is this a serious question? By buying any option you are going long on volatility. How you go about it is a completely different matter. You can be right about vol going up and still lose plenty of money. It is a lot more complicated to execute effectively (other than just blind luck) than by buying straddles on the SPX or calls on the VIX. 2) The only real way to profit off of a long volatility position is either by selling out your position when vol catches an extreme bid (ie flash crash) or through gamma, which would then require you to lock in the deltas at some point anyway--a big, long rally or sell-off (which will likely be accompanied by a period of higher implied vol) is likely to cause extreme losses (far, far more than your initial investment) for you position if you are constantly delta hedging. Hedging frequency is a complicated matter and has been discussed ad infinitum with no great conclusions (there is little difference in EV and it mostly gives smoother outcomes for your M2M rather than actually providing some great benefit). As a retail investor, it is mostly worthless to think about because transaction costs will kill anything you could possibly make and you won't have a big enough position anyway to be gaining/losing that many deltas (how many options do you really plan on buying?). 3) ???? There is no 'investing' in the VIX. I am not even talking about the lack of a real futures market--I am talking about the fact that it is not an investment in the first place, but a pure speculation that there will be an increase in volatility. With the way the VIX is computed, this might not even replicate what you want it to in the first place.
Yes, this is a serious question. I understand that you gain exposure to volatility by buying an option, but what I was asking was about constructing a portfolio that had a high degree of correlation to the VIX and low correlation to the price movement of the S&P500 (i.e., a synthetic VIX).
I'm trying to understand how this would be done from an institutional perspective, where presumably transaction costs could be kept low relative to any profit or loss. Can you suggest any articles on hedghing frequency?
I understand that you can't invest in the VIX, hence me asking the question in the first place. What I'm trying to understand is how close the replicating portfolio would perform against the VIX
If there are big moves, this is great for you being long gamma, especially if you are anticipating this and not just hedging delta based on a static set of rules, e.g. delta >2m or every other day or so. Being long gamma is the very essence of being long vol. You are becoming delta longer when the market is going up and vice versa, that's why you are making money. If there are no such moves, realized vol is lower than implied vol, and therefore you will loose. So the profit is coming from how you delta hedge.
There is no other way of being long vol if you do not want to hedge delta e.g. than with vol or var swaps (besides ETFs or other structured products).
If you are using options, you are not only able to bet on vol, but also on the shape of the vol smile: Straddles for the delta neutral vol level, so you buy options whose strikes are ATM. Risk reversals, so a shift in the curve, at the moment you could bet on the EURUSD skew coming down with a .25 delta put short and .25 delta call long. The steepness of the vol smile by buying a .10d butterfly, so you are expecting a higher implied vol to be priced in at the very outer part of the smile. You think extreme moves in the underlying are more likely than the implied vol would predict.
Straddles and strangles for short term volatility bets.
Also, just so we're clear, you can't possibly 'invest' in the vix or volatility even. There are no returns on volatility of the S&P and it is mean reverting. Even if it is just a speculative bet that volatility will be greater than implied, that has historically been a massive sucker bet. The premium you are going to pay out to going through the B/A spread is going to clean you out typically.
Not a single thing you said relates to simply going long volatility like the OP is talking about, but rather trading parts of the skew.
I understand you are getting long/short delta with long gamma--that isn't exactly an earth shattering fact. The problem comes when you are frequently delta hedging and paying large premiums. Extremely sudden large moves will be fine long gamma, but simple large moves over a significant period of time are not enough because you will be constantly hedging throughout, cutting yourself up in the process and not earning back the premium. Not delta hedging regularly is all well and good in theory, but you will be fucked if it is range bound or you still pay too much of a premium. It also makes for an extremely bad M2M situation where you will have on tons of deltas one direction or the other and have problems exiting the position easily depending on the market.
Obviously you can trade based on shape of the smile, but that has nothing to do with what the thread is about. BTW you may want to actually see what happens with RRs and butterflies. All of that is well good in theory, but in practice people aren't using RRs as a volatility bet, but rather as a way to put on a directional bet. People aren't trading butterflies because they believe the wings are going to be bid/offered....
Maybe you should read my post again.
I am saying that there is no simple solution of going long vol, because you have to know which part of the smile you want to go long. Premium does play a very small role in my world (FX options), and if you are trading S&P mini futures or sth. similar, I guess you have a small spread as well.
I am on an fx options desk, and we are trading with straddles, strangles, risk reversals, butterflies etc. People are using these strategies for directional bets usually, that's correct, because they don't bet on pure vol., but that's what the OP wants.
In essence: There is no easy way of going long vol other than vol or var swaps, leaving structured products aside. With options you will always have the path dependence problem and high correlation to the movements of the underlying.
Iusto soluta quaerat praesentium odit. Dolore non totam atque dolorem molestiae totam et. Quo nesciunt totam unde. Consequatur minus sit voluptates exercitationem temporibus. Ipsam cum voluptatem in iste assumenda facilis non.
Illum at unde sapiente in. Esse ut quis cumque sed incidunt dolor. Omnis quos quia tempora aut et.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...
Autem consequuntur voluptatem quo eaque. Voluptates nemo corporis et voluptatem praesentium sapiente beatae. Laudantium non corrupti et inventore.