Member D. S. posed the following question:
I’ve been trading SPY iron condors for some time now and I have been opening them at very similar levels to yourselves. However I have been using much wider spreads, so whereas you use a $2 spread I would use as much as a $10 dollar spread. What in your view is the value in only using smaller spreads? My reasoning on the larger spreads is as follows: 1) You can open the trade for a larger amount and therefore break-even is further out than with smaller spreads 2) While there is more money at risk, the risk level is much lower. e.g. if underlying crosses the lower strike by $2 in your case max loss is realised whereas in my case only about 20% of max loss is realised.
For the newer traders out there, let’s review the different approaches he’s talking about. One way to think about an iron condor is as the sum of a short strangle and a long strangle. So the construction of an iron condor involves two steps: in the first step, we want to sell a short out of the money strangle, and the primary decision here is which strikes to choose, i.e., how far out of the money should we go? That’s an important question in itself, but it’s not the topic here. The second step is to buy a long further out of the money strangle in order to hedge the risk in our short position. In most cases we want the long call and long put to be equidistant from the short call and short put that they’re hedging; but, as D.S. identifies, it’s not necessary that the long strangle be one strike outside the short strangle.
On first glance, it really doesn’t matter where those long strikes are: if you want to risk a particular amount on the trade, it doesn’t matter whether the distance between your short and long calls (or short and long puts) is $2 or $10, provided you adjust the number of contracts traded accordingly. For example, with IWM at 51.61, let’s say we wanted to trade an iron condor with short strikes at 46 and 56, meaning we’re selling the 46 puts and selling the 56 calls. Now, if we were opting for a one-strike range, we’d obviously buy the 45 puts and the 57 calls; if, as D.S. suggests, we wanted to use a ten point range, we’d buy the 36 puts and the 66 calls. Here’s what those two trades might look like:
+1 IWM 45 put +1 IWM 36 put
-1 IWM 46 put -1 IWM 46 put
-1 IWM 56 call -1 IWM 56 call
+1 IWM 57 call +1 IWM 66 call
$0.29 credit $0.98 credit
Assume we want to risk $5000 on this trade. In the first variation, the maximum possible loss on a 1-lot trade would be $71, or the distance between the short and long options ($100) less the credit received ($29). To risk the desired amount, we’d trade 70 contracts of the first position. In the second, wider variation, the maximum possible loss on a 1-lot trade would be $902 ($1000 – $98), so to risk the desired amount we would trade 5 contracts. (Since that second variation is actually only risking $4510, I’ll set the first variation to use 63 contracts so that they really are risking the same dollar amount.) To address D.S.’s first point, while it’s true that the second variation brings in a larger credit up front, and while the break-even points at expiration on the second trade are slightly further out, I don’t regard this as a significant factor, for a few reasons. First, I rarely hold trades through to expiration. Secondly, assuming we risk the same amount on each variation, any differences in the nominal credit received will obviously balance out. Finally, the conditions that cause me to exit and enter trades have nothing to do with some fixed profit/loss level, but have everything to do with the various greek exposures I want to maintain.
The second claim made in favor of the wider variation is that it is a lower risk trade. I don’t see the logic behind this claim. In the first place, we’re risking the same amount on each variation, and the probability that the underlying will be between the strike prices of our short strangle isn’t affected by which long options we buy. The comparison of maximum loss points relative to a move in the underlying appears to be made assuming we’re at or near expiration; of course, the real profits and losses incurred between now and expiration will depend on price changes, but also on changes in implied volatility.
And that’s where this wider approach to the long strangle looks less desirable. Remember that the point of an iron condor is to sell options that you believe are priced with an implied volatility that is higher than the realized volatility you expect in the future. The table below (click to enlarge) displays the option chain for the July 2009 IWM options that we’re discussing. At the time of writing, the 46 puts and 56 July calls we’re selling have implied volatilities of 42.74% and 28.93% respectively. If we buy hedges that are nearby – say, one or two strikes away from the options we sold – the implied volatilities of those hedges won’t be much different. But look at the implied vols ten strikes away: both the 66 calls and the 36 puts have significantly higher implied volatilities than the options we are short. That’s called vertical volatility skew, and it’s a completely ordinary feature of options markets.
But if the point of this iron condor spread is to sell “overpriced” volatility, we should clearly try to pay as little as possible (in terms of implied vol) for the hedges we buy; which means that we’re better off paying for those nearby puts and calls, rather than for the distant long strangle. Put another way: it’s likely that we will always give up a bit of edge insofar as we’re selling IV at one level and buying it at a slightly higher level. But while the penalty there is small – and certainly not significant enough to warn us off the strategy – it also doesn’t make sense to exacerbate the situation needlessly. Overpaying for your hedges isn’t a problem that will show itself in any dramatic way, but it will steadily erode returns over time.
Traders who overpay for hedge positions because they want to reduce the size of their commission bill are merely swapping a transparent cost for an opaque one.