Last week, we again engaged in a little play therapy. We concentrated on a November SPX iron condor theoretically established on Thursday, October 6, the day before the non-farm payrolls had been released. In that article we had talked about various ways we might insure or hedge against adverse price action to the downside. Declines, particularly sharp ones, often result in increasing implied volatilities. Although it can be helpful to hedge once a downturn has begun, that hedge will often be more expensive than it would be before the downturn occurred. For that reason, I tend to always buy an extra long put on my iron condors when the trade is established.
I don't often buy such insurance for the upside. I tend to wait until I need it. Why is that? Often--although not always--when prices rise, implied volatilities drop. For example, the aggregate implied volatilities of the SPX NOV options was 38.55 percent when that trade was first established. By Friday, October 14, 2011, it had dropped to 35.96. That decreases the value of a long call, working in opposition to any increase due to price action. So does the passage of time, of course, and eight days had passed. These factors result in the sad reality that the long call bought at the inception of a trade might not always be as helpful as you thought it would be when it is needed.
A happier reality is that it might not be needed as soon as it appeared it would be needed from the "today" graph at the inception of the trade. Let's borrow the original "today" profit-loss line for the iron condor from last week's article. That iron condor had sold strikes at 955 and 1295, with the long options of the credit spreads 10 points further out of the money.
Projected Profit-Loss Line for Friday, October 7, from Last Week's Article:
Follow that blue line across to 1220, and you can see that the loss projected for a sharp rise, if had happened that Friday, would have been around $2,000.
If nothing had been touched on that iron condor, what would the theoretical loss have been on Friday, October 14? Note: this is "theoretical" because I didn't trade this iron condor and so of course cannot guarantee that I could have either entered or closed the trades for the amounts projected on these graphs.
Theoretical Profit-Loss Line for the Same Position, on October 14:
Please note that, since I always add put insurance, this graph does include the put insurance. As a result, the left side of the graph does not slope down as steeply as the right side. The red dot is positioned at the SPX's price near the close on October 14. Eyeballing this chart reveals that the loss is less than originally projected for this price point a week earlier. The passage of time and the lowering of implied volatilities has helped, as has the lowering of values in the put credit spread side. The theoretical loss is about $98.00, a figure calculated by Options Oracle and not visible on this chart due to sizing requirements for publication here.
Imagine that we had employed the same tactic for upside protection as I typically use for downside protection: spending 10-15 percent of the credit taken in to buy a long call. Last week, I imagined spending about $175 to buy an extra put on the Thursday, October 6, that the trade was established. If I had decided that there was more risk of an upside breakout than a downside breakdown and I didn't know what I know about upside protection, I might have elected to spend a similar amount on a long call instead of a long put. If I had spent a similar amount on buying an extra long call, that likely would have meant buying one NOV 1335 call for $1.70. What was that call worth on Wednesday, October 12, when the SPX first approached the 1220 area? How much had it gained in value? One source listed it at $1.15.
It had not gained any value. It had lost. It offered no protection against a rally, even a hard rally over a short period of time.
In fact, I've heard one frequent iron condor trader suggest that a call won't be helpful unless it's inside the sold strike, which would make it far more expensive than the 10-15 percent I typically spend on protection.
Some traders employ debit spreads in front of their credit spreads. Because one option is sold and another bought, the effects of changing volatilities is somewhat eased. If one is expensive due to inflated implied volatilities, the other is likely to be, too. Let's imagine that on Thursday, October 6, when this theoretical position was established, an ITM 1150/1160 debit spread was established. It could have been bought for $5.80. What would it have been worth on Wednesday, October 12? One source quoted it at $7.25, a $1.45 x 100 = $145 gain, minus commissions. That means that employing this method would have reduced the $98 theoretical loss by $145, minus commissions, resulting in a gain of $47.00. It wasn't a huge help, but it certainly helped more than the purchase of a single long further-out option.
Of course, it cost far more, too. The trader who employed this debit-spread method might have wanted to have a plan in place for removing the debit spread if markets moved too far to the south or else be willing to accept a far reduced profit.
Take another look at that second graph. Notice that the red dot seems poised at the edge of a steep slide down the profit-loss graph? The SPX had zoomed up a seemingly impossible amount in a little more than week--although we now know that move wasn't nearly over--and it was at this time approximately 70 points away from the sold strike. At the time, we might have reasoned that it seemed unlikely that the SPX could maintain such a steep rally without pausing to digest gains, but what if it didn't? What would the graph look like if a debit spread were purchased then? I was writing this on October 14, so all these suppositions were made and graphs snapped at that time.
Let's try an ATM one. Due to the presence of the insurance put and the internal debit spread, the expiration graph has begun to look a little lumpy.
New Expiration Graph as of Friday, October 14, 2011:
New "Today" Profit and Loss as of Friday, October 14, 2011:
Although the sizing requirements for publishing these graphs may make it difficult to see the differences, this graph falls off a little less steeply than the previous "today" graph, also reflected in the smaller absolute value for the delta. It's going to be hurt a little less by price action to the upside, but we may not be wowed by this result.
What if we widen the spread to twenty points? Wait a minute, though. That's going to cost us a hefty amount. If we do that, we have to have a plan for taking profit in the debit spread and maybe simultaneously closing out some of the credit spread contracts at a loss, thereby reducing the risk. We also probably need a plan for the way we would treat that expensive debit spread if the markets turned south. Luckily, we would likely have a little time to enact that plan because a south turn would likely pump up implied volatilities again and keep the debit spread from losing too much money too quickly.
New "Today" Profit and Loss Graph with a 20-Point Debit Spread:
This one gives more upside protection but the curve steepens now to the downside. This reflects the fact that this adjustment changes the overall position delta from its pre-adjustment negative value of -21.23 deltas to a flattish -3.41 one. Do we really want to do that after the SPX has been on such a tear? What is our view of what will happen next? I usually like to keep my iron condors and butterflies a little negative delta because they're going to be hurt if markets turn south and implied volatilities pick up. They're negative vega trades. So, how much of a change in the delta do we want? That depends on our view of what will happen next in the markets. Some traders like to keep that delta as flat as possible, but they'll pay for that. Others like to reduce the delta risk by a third, a half or two thirds, depending on their own risk tolerance. This adjustment brings down the possible profit on this position to $855.00 from the original $1,620 after the put purchase, even if all else goes well.
Hmm. What else could we do? What if we bought a 1280 to go in front of our sold 1290's, and sold one of the original 1300 longs, effectively putting a debit spread right in front of the credit spread? That brings the delta from the pre-adjustment value of -21.23 to a less negative -15.29, with the possible profit decreasing only to $1,435.00. At the size at which I can include the graphs, this one won't look much different than the pre-adjustment one, the first "today" graph I included. For that reason, I won't post it here, but the changed delta values show that it will have slightly lessened the negative impact of a further rally. It also slightly reduced the positive theta, however, from 49.08 to 44.84.
What else is possible? What if we just bought in one of the sold 1290 calls? That would have cost $4.00, reducing the overall profit potential to $1,220, but it brings the negative delta from -21.27 to -7.30. It also reduces the theta quite a bit, however, to 30.87. It's a personal preference, but I often like this adjustment because it reduces risk. This month, for example, I bought a debit spread when I was fairly certain the rally was going to stick for at least a few days, took profit in it at a certain point, and used that profit to partially offset the cost of buying in a sold strike.
If you haven't previously used a profit/loss graph system like this or haven't studied the strategy's Greeks, this may all look confusing. However, all we're doing is asking some "What if?" questions, inputting the ideas that are proposed and looking at how it impacts the graph over the next day or two. We can look a bit into the future and examine several scenarios (markets up with no change in implied volatility or markets up with a lowering of implied volatility, markets down with a rise in implied volatility) and determine how soon our positions will get into trouble. We can try out the effectiveness of several adjustments, realizing that each will have a downside as well as an upside. Those new to using graphing programs such as these should realize, however, that the "theoretical" graphs don't always play out the way they're shown in actual trading. The skew of options can change, so that the implied volatility steepens more in puts or turns up in both calls and puts.
Still, it's important to keep the idea of "play" in focus when you're looking at these and thinking about the different kinds of adjustments that might be possible. Doing so let's us test idea that might seem dumb, perhaps discovering something that's more workable than we might have first thought. It lets us experiment freely.
I used freeware Options Oracle for a while. I now use a proprietary graphing system that's far more flexible and allows for extensive back testing, but which is unfortunately not yet available to the general public. Others exist, however. Spend some time playing around with such adjustments and thinking about them.