Verticals - OTM vs ATM vs ITM (time decay and volatility decrease)

[ehsuhuang]

Can you guys verify if my thinking is along the correct lines on vertical spreads?

My general strategy: is to swing trade equities, and hold them for about 3 to 6 weeks as an average timeframe. I am usually correct about 75% of the time, within the 3-6 week timeframe. A typical trade might be to buy SPY today, with an exit at 110, and stop at 70. My position size would be to win or lose 5% (if SPY hits my stop at 70, I would lose 5% of my trading account).

Options: I started to buy call options on the underlying equities, as a substitute to reduce overall risk. I had a couple of occasions over the years where the equity would gap down by 20-30%, and I didn't want to expose that much of my account. In the SPY example above, I would usually buy the calls about 4-6 months out to allow enough time.

Since buying calls, my overall performance has not been nearly as profitable. While my win/loss rate on the underlying stock / strategy was about the same. However, when my trade went well, I would only make a 3-4% profit. And if it did not go well, I would lose 6-7%. These were due to the time decay over 4-6 weeks, and usually a small drop in implied volatility (majority was due to time decay).

Now trying Vertical Spreads: So, I wanted a simple strategy with limited risk, limited profit (I always exit at a predetermined point anyway), little risk for early assignment (I don't really want to start selling a lot of options), and relatively neutral to time decay and implied volatility decreases. I am leaning towards doing Long Call Verticals (or Long Put verticals if I think the stock is going down).

My question is: given my underlying stock swing trade methodology described above, is it more optimal to buy OTM vs ATM vs ITM vertical spreads? Playing around on Think or Swim's Analysis page, it seems to me that the optimal strategy would be to buy a slightly ITM call, and sell the OTM call. Thus in the SPY example, I would buy the June 85 calls and sell the June 110 calls. The value of the spread should go up slightly with time decay, and also go up slightly when implied volatility drops a little.

Is my thinking on the correct path here?
Thanks for the help!

 

[dmo]

If you want to play the direction of the S&P 500 with options - and this is true of SPY options, SPX options, and ES options - it's important to keep in mind that every time the S&P 500 goes up, IV will drop. And every time the S&P 500 goes down, IV will rise.

There is no more consistent phenomenon in all of trading. So you need to make it work for you, not against you.

That may be why you found it difficult to make money buying calls. Each time the S&P went your way (UP), IV dropped and you made less money than you "should have" made.

You're on the right track with your plan to buy an ITM call and sell an OTM call. That way the IV pattern described above is working for you.

Why? Because if SPY is at 90, and you buy an 88 call and sell a 92 call, the calls are equidistant from the money and thus have approximately the same vega. So overall you are approximately vega neutral.

But if SPY goes down - say to 88 - now your long 88 calls will have much more vega than your short 92 calls. Overall you will be long vegas. And - conveniently - IV will rise, counterbalancing your loss from being long deltas. Nice!

And if SPY goes up, say, to 92? Now your short 92 calls will have far more vega than your long 88 calls. Overall you will be short vega. And lo and behold, IV will go down, increasing your profit from being long deltas. Nice again!

 

[Mr.Consistent]

Quote from ehsuhuang:
Is my thinking on the correct path here?
Thanks for the help! [/B]

Do a google search for "morning market huddle" by Steve Lentz, you can view his archives, he mainly does verticals and occasionally bullish butters and bearish calendars.

BTW; If you're good at swing trading equities, try demo trading Forex, in FX you don't get nasty surprises as in stocks which means you can more less predetermine you ultimate risk as you can with options, plus there are nice trends, and plenty of leverage. Why muck around with spreads and limit your profit potential when you catch a trend? Just a thought

All the best

 

[ehsuhuang]

Mr. Consistent - funny you mentioned forex. I traded that about 5-6 years ago when all the internet broker/bucketshops just came out. It was good for all the reasons you mentioned, but it since there are only 5-6 major currencies, the patterns that I was looking for didn't occur frequently enough. I tried daytrading it for awhile, and remember staying up really really late for the London open at 2 a.m. Needless to say, that schedule only lasted for a couple of weeks.

DMO - thanks for the info, and I agree with volatility decreases hurting my strategy. I estimate that the volatility drops have reduced my overall profits by about 10-20%. (And the time-decay reduced it by another 20%). On further analysis, I found a flaw in my verticals plan of buying the slightly ITM call and selling the OTM call. If my underlying trade works out and the SPY goes up from 88 to 100 in the example above, the vertical spread will protect my theta and vega. However, when my trade doesn't work out as expected, and the SPY drops to my stop of 70, both calls are now way OTM and my vertical spread still becomes vulnerable to time decay and volatility decreases.

To put it another way, using an ATM vertical spread achieves my goal of protecting theta and vega for the 75% of my trades that work out. But for the other 25% of my trades that don't work out, the spread is still vulnerable to time decay and volatility just like simply buying a call outright (although slightly less so).

I've though of buying the vertical spread with both calls deep ITM, but the spreads on most stocks are too wide, and the profit factor versus capital at risk becomes much smaller.

 

[nitro]

Quote from dmo:

...it's important to keep in mind that every time the S&P 500 goes up, IV will drop. And every time the S&P 500 goes down, IV will rise.

...

Please stop saying every time. That is simply incorrect.

 

[MasterAtWork]

Hi,

Happy new year.

Nitro, you're right but Dmo just can' t stop.
As a former market maker, it's really weird that he never took a look at volatilities for long term options and their behaviours.
If you read some of his posts, he kept focusing on VIX variations and take it as a given for all.
Course it's wrong and everyone who takes the time to look at volatiliIES even on S&P would support this fact. VolatilitIES don't shift the same way, don't react the same time, the same direction.
It would be so easy to make money for market makers if volty markets would behave like that. No research needed and a lot of easy money would be made. Forward volty would be easy (easier) to forcast.

Do you remember Dmo, interest rates CAN'T be negative as you stated. So a delta CAN'T be higher than 1 for vanilla options. And I never read one of your post where you finally agreed it. But reality did.

But Dmo got the right words to be understood by newbies, and this way his contribution is great.

 

[dmo]

Quote from nitro:

Please stop saying every time. That is simply incorrect.

Nitro - haven't we been through this before? Not once but many times?

The pattern seems to go like this: I say that every time the S&P 500 goes up, IV goes down, and every time the S&P 500 goes down, IV goes up. You come in and say that's not true. I post a chart of the last few days demonstrating that it was true for those days at least, and invite you to post a chart showing otherwise. So far you haven't taken me up on it.

Just so as not to break with tradition, I'm posting a one-minute chart comparing the VIX with the SPX, with the SPX on top, VIX on the bottom. Maybe you can find a few minutes at a time here and there where the inverse correlation does not hold but overall, the pattern is obvious and reliable.

For an individual option (as opposed to the VIX as a whole), it's true that there is a phenomenon that acts counter to the correlation mentioned above. Namely, the skew, which in options on the S&P 500 is ALWAYS (yes, always) such that the lower the strike the higher the IV, and the higher the strike the lower the IV. So if we're talking about a 920 call for example, then as the S&P 500 rises, the overall volatility as measured by the VIX will go down. However, that 920 call will now have a "more favored" place in the skew, and so its IV will not go down as much as the VIX would seem to predict.

I think that's what MAW is very correctly referring to.

Still, overall, even considering individual options and not just the VIX as a whole, I think you will find that the overall negative correlation between the SPX and IV is very valid.

 

[dmo]

Quote from MasterAtWork:

Hi,

Happy new year.

Nitro, you're right but Dmo just can' t stop.
As a former market maker, it's really weird that he never took a look at volatilities for long term options and their behaviours.
If you read some of his posts, he kept focusing on VIX variations and take it as a given for all.
Course it's wrong and everyone who takes the time to look at volatiliIES even on S&P would support this fact. VolatilitIES don't shift the same way, don't react the same time, the same direction.
It would be so easy to make money for market makers if volty markets would behave like that. No research needed and a lot of easy money would be made. Forward volty would be easy (easier) to forcast.

Do you remember Dmo, interest rates CAN'T be negative as you stated. So a delta CAN'T be higher than 1 for vanilla options. And I never read one of your post where you finally agreed it. But reality did.

But Dmo got the right words to be understood by newbies, and this way his contribution is great.

Hi MAW - I knew you'd show up when I said "always."

MAW, some time ago you explained how, in your opinion, an option's delta could be > 1.0. I never said that wasn't true - just that I couldn't figure out your reasoning. That's still true. So you can chalk that one up to stupidity on my part, not obstinacy.

As for my conviction about the negative correlation between SPX and IV - now THAT you can chalk up to obstinacy! Why don't traders then make a fortune if it's so reliable? As I explained in pretty intimate detail in a thread a few months ago, the existence of the skew takes that easy play away. The reliability of the skew pattern - where the lower the strike the higher the IV - is a direct reflection of the negative correlation between the SPX and IV.

 

[MasterAtWork]

Hi Dmo,

Quote from dmo:

Hi MAW - I knew you'd show up when I said "always."

MAW, some time ago you explained how, in your opinion, an option's delta could be > 1.0. I never said that wasn't true - just that I couldn't figure out your reasoning. That's still true. So you can chalk that one up to stupidity on my part, not obstinacy.

As for my conviction about the negative correlation between SPX and IV - now THAT you can chalk up to obstinacy! Why don't traders then make a fortune if it's so reliable? As I explained in pretty intimate detail in a thread a few months ago, the existence of the skew takes that easy play away. The reliability of the skew pattern - where the lower the strike the higher the IV - is a direct reflection of the negative correlation between the SPX and IV.

The reasoning is very simple to handle, but rarely shown and explained. So it's for you because the new year

If negative interest rates are hard to imagine (but last months showed they exist, like negative swap spread to Treasuries), one would prefer a global explanation with cost of carry.
The cost of carry is defined as: b=r+s-d
where r is interest rate, s is storage cost and d is either dividend or foreign interest rate for currency options.

Imagine you're pricing a simple call on commodity. You need to adjust your model with the cost of carry. That is the storage cost of commodities. This means that just holding the commodity will result in a gradual loss of wealth even if the price remains fixed. This is like having a "negative dividend". And a "negative dividend effet" on option pricing leads to a delta that is >1 as you know.

Delta call is =exp(b-r)*(N(d1)) in Black and Scholes world
and b-r=(r+s-d)-r=s-d
If there is no dividend but storage cost, d=0 s>0 and exp(s-d)=exp(s)>1
For a deep in the money call, N(d1)=1
thus Delta Call=exp(s)*N(d1)>1

The same if there is no storage cost but negative "dividend" or negative foreign interest rates for currencies .

I will let you find which kind of "storage cost" one could find on different markets.


For the fact of skew existence, as I told you there are a lot of reasons which can explain the skew.
One of them is that basically, if you work as an asset manager, your boss will likely tell you to buy puts, not calls. So demand and supply rules will do the rest of the job.

About ALWAYS negative correlation between IV and SPX, please try to take a look at what is going on with S&P DEC10 and DEC11 implied volatility options. I'm sure your obstinacy won't last .

 

[dmo]

Quote from MasterAtWork:

Hi Dmo,



The reasoning is very simple to handle, but rarely shown and explained. So it's for you because the new year

If negative interest rates are hard to imagine (but last months showed they exist, like negative swap spread to Treasuries), one would prefer a global explanation with cost of carry.
The cost of carry is defined as: b=r+s-d
where r is interest rate, s is storage cost and d is either dividend or foreign interest rate for currency options.

Imagine you're pricing a simple call on commodity. You need to adjust your model with the cost of carry. That is the storage cost of commodities. This means that just holding the commodity will result in a gradual loss of wealth even if the price remains fixed. This is like having a "negative dividend". And a "negative dividend effet" on option pricing leads to a delta that is >1 as you know.

Delta call is =exp(b-r)*(N(d1)) in Black and Scholes world
and b-r=(r+s-d)-r=s-d
If there is no dividend but storage cost, d=0 s>0 and exp(s-d)=exp(s)>1
For a deep in the money call, N(d1)=1
thus Delta Call=exp(s)*N(d1)>1

The same if there is no storage cost but negative "dividend" or negative foreign interest rates for currencies .

I will let you find which kind of "storage cost" one could find on different markets.


For the fact of skew existence, as I told you there are a lot of reasons which can explain the skew.
One of them is that basically, if you work as an asset manager, your boss will likely tell you to buy puts, not calls. So demand and supply rules will do the rest of the job.

About ALWAYS negative correlation between IV and SPX, please try to take a look at what is going on with S&P DEC10 and DEC11 implied volatility options. I'm sure your obstinacy won't last .

You're a math guy maw, and I'm a nuts and bolts practical type. But I can see how the two can meet.

To me, the test of true delta is this - if I'm long 100 very, very DITM calls, how many futures do I need to sell to be really neutral?

If interest rates are 10% and I'm long 100 of the US T-bond future 100 calls, and the futures are at 140, and I'm short 100 futures, then I'm not delta neutral. Why? Because every point the futures go up I have to come up with $100,000 in variation margin. My clearing firm will be happy to lend me that money as it's a safe bet for them, but I'll have to pay interest on it. So in reality I'll have to be short somewhat less than 100 futures to compensate - meaning the true delta of those DITM calls is < 1.0

Now, let's look at a practical application of your negative interest rate scenario. If my clearing firm truly was willing to pay me interest on the money they lend me - to compensate me for taking money off their hands, that would be a true negative interest rate in the real world that I inhabit. And if I had to pay interest on the variation margin I take in when the futures drop (since I'm short futures), that would also be true negative interest rates.

So yes, in the case of true negative real-world interest rates, I can see how delta would be > 1. At the same time, in my life, I can't imagine how that would happen. I'll concede though that there must be special circumstances between institutions where it could.

As for the skew, haven't I been repeating like a mantra for the past year that it's the result of supply and demand, driven by the need for portfolio insurance, rather than the market implying some sort of probability distribution as commonly thought?