• Nemo’s Bond Crash Course – One of my resolutions for 2013 was to understand fixed income better. His recent bitcoin explainer reminded me of this bond intro. Also: @groditi, David Schawel
• Goldman: AUD/USD has plenty more downside – I’ve been buying straddles and puts here and there since late last year (viz. early = wrong). The recent breakdown in AUD has been validating and it looks like foreign direct investment is moderating.
• Hendry’s Eclectica are long Japan – “Back in 2008, we purchased a ten year 40,000 Nikkei one-touch call option.” Huh.
• Stealing from the internet – it’s about how to scrape web data into R.
• JGBs: “Yes I would, Kent” – no reason to worry quite yet about Japan. Simpsons clip is just right.
• Pair trading index options
• Nobody is committing to Jumbo SPY options quite yet – maybe we have enough S&P 500 products for now?
• There is zero correlation between the Fed printing and the money supply. Deal with it. – Very good.
• And finally, some hard-core metaphysics by me: Material Composition and Theory Choice and the tl;dr version.

On May 6, 1994 Robert Francis Goldthwait made his mark on television history.  Better known as Bobcat Goldthwait, he chose for some strange reason to set fire to the interview chair during a guest appearance on the Tonight Show.  This incident lasted only 30 seconds or so as Jay Leno used his beverage to douse the flames.  I think what Bobcat was trying to do, in his own unique way, was to demonstrate that volatility spikes are often doused very quickly in the financial markets.  I commend him for this bold lesson in not panicking when volatility hits the stock market. [CBOE Options Hub]

The lesson is that when, from time to time, some crazy incident spooks the market, there are usually adults around – in the form of cash-rich value investors, central banks, or other agents – to calm things down. As a Letterman/Conan fan, I draw an additional lesson from the analogy, which is that even subpar market actors can get the job done, provided they have the right tools.

And finally, this scene from the Larry Sanders show teaches us that volatility comes in clusters.

The Condor Options newsletter

The purpose of the strategy is to profit from the volatility risk premia that are priced into options. While the volatility risk premium is a consistently observable, ongoing feature of option markets, the strategy does not adopt an agnostic, “always on” approach. Instead, we select trades in markets where the expected risk-adjusted premium is the greatest.

One of the most frequently asked questions we receive is how to manage positions that are under pressure or are unprofitable, and there have always been vendors and others selling tips on how to “adjust” losing iron condor positions; one unique feature of our strategy is that it does not depend on overwrought, complex adjustment techniques, relying instead on straightforward delta and gamma updates and the intuitive idea of spreading risk across a book of trades. Real-time and historical results speak to the effectiveness of our approach.

Performance data for the Condor Options newsletter is updated at the Performance page, including a trade sheet and monthly return data.

Our average hold time for each position has been about 35 days, and our delta hedging trades are updated weekly but only when needed – so this isn’t a strategy that requires constant attention or active management. We are accepting new clients at this time.

@condoroptions My interpretation is more benign. I think that the market is just pricing lower average volatility long term.

— Eli Mintz (@VixCentral) April 30, 2013

Here’s why this is a plausible explanation of those recent flattening periods in the VX slope. If the market is expecting lower volatility over the long term, that should make VIX futures prices at longer horizons more “sticky” during short term volatility spikes. So if SPX declines 3% or so but lower long-term vol expectations are in place, we would see some movement at the front end of the VIX curve but not as much on the back end, and that would entail some flattening in the term structure even if the short-term vol reaction wasn’t extreme.

That’s a persuasive idea and it would make sense if it were true. How much weight should we give the notion of lower vol expectations? This chart shows the daily changes in VX June futures over the last few years – excluding April, May, and June since the futures become much more volatile closer to expiration. Over the last few months, the June 2013 contract has been acting normally: less volatile perhaps than the June 2012 contract was, but the same or more volatile than the 2010 and 2011 contracts. The mean value of the most recent series is lower than in previous years, which is also consistent with the idea of a gradually lower price for medium term volatility.

At the same time, there’s still something to the idea that markets have been more reactive lately in the short part of the curve. Look at the ten day standard deviation of the daily VIX close:

The most recent spike was the highest we’ve seen since the August 2011 European debt crisis. And you can clearly see the three spikes in this series corresponding to the flattening term structure that inspired my earlier post. Another way to characterize this is that the realized volatility of implied volatility is higher than it has been in some time.

Since December 2012, however, changes in the slope of the VIX term structure have been more violent than in years past. The attached chart shows the log ratio of the maximum and minimum VIX futures values on each day (blue) along with the S&P 500 over the same period (black). Since we haven’t seen many periods of sustained VIX backwardation in recent years, it’s safe to read the Y axis such that high values indicate a normal contangoed market and low values show a flattening term structure (i.e. market stress).

Prior years have been characterized by rough but discernible trends. However, since the fiscal cliff situation captured the attention of investors in late 2012, there have been three occasions in which this slope estimate swung from 0.40 to 0.15 or lower. Again, those sharp declines indicate a flattening term structure, with short-term implied volatility rising faster than longer-dated option premium.

The speed of those declines and snap-back resolution shows a market that is quicker to react negatively than we’ve seen in recent years, even with asset prices at multi-year highs. Before December, the sharpest flattening in the term structure was in May, when the market saw a drawdown of more than 9%; the most recent bear flattening needed a market decline of just 3.8%. Investors could probably stand to relax a little. At the very least, this indicator shows the very opposite of complacency.

Fig. 1. S&P 500 Index, March 6 – April 18. Source: CBOE

Now, trading strategies can be classified according to the source of their risks/returns, meaning that we can sort strategies into the buckets known as option greeks. For example, a buy and hold investor is long deltas; a mean reversion strategy is short gamma. What sort of risk exposure worked well recently?

1. Owning deltas means profiting from a directional move in the underlying. Owning deltas, whether positive or negative, was not a winner over this period. The market made a round trip back to 1,540.

2. Owning gamma (by buying calls or puts, or both) means profiting from convexity. It’s analogous to a trend-following strategy in which you commit more capital as momentum builds. That didn’t work well, either: the carrying costs in early March were recovered on the trip to 1590, but then given up (and then some) when the market reversed.

3. Theta is the cost of owning that convexity. Being long theta / short gamma worked in March and we’ll come back to this.

4. Owning vega means profiting when implied volatility in an option rises. This is where a lot of people will put up a chart of VIX and say: see, implied volatility exploded, VIX moved from 13 to 18! But that ignores the roll down costs associated with a real product. Instead, look at how May VIX futures traded over this period: they fell from 16.50 to 14, and then rallied back above 17. That means being long vega saw a drawdown during the period more than twice the size of the eventual gain. If you closed out a long May VIX position on Friday, you did so profitably but only after some rough trading.

Fig. 2. May VIX futures, March 6 – April 18. Source: CFE

The best strategy over this period was to be short gamma / long theta. The best way to adopt that sort of exposure is by trading iron condors. There are other ways to be short gamma – like selling straddles or strangles naked – but that’s a riskier and more difficult approach; a condor is essentially a short strangle covered by a long strangle. March and April were a perfect demonstration of how gamma-oriented trades can augment portfolios consisting of other strategies and investments. If the market makes a round trip, price-wise, you won’t make anything by owning shares, and if implied volatility is flat or higher, short volatility trades will struggle to pay their way. But no matter what else happens, time still passes, and in this sort of market, it pays to be on the right side of time.

Fig. 1. CVIX – Major FX 3M IV. Source: Deutsche Bank

The May GLD strangle I recommended earlier this year paid off well, although to be honest I wasn’t expecting anything quite this dramatic. Active traders should actually favor the short vol side here. Look at the spike in one month GLD implied volatility (GVZ) as of Friday:

Fig 1. GLD and GVZ, 2012-2013. Source: CBOE, Condor Options

At pixel time, GVZ has vaulted up again, near 32. There’s no word yet on changes to COMEX margin requirements, although it’s safe to assume those changes are coming. You can monitor those requirements here. To the usual reasons for favoring short volatility positions after a spike, add the fact that higher margin requirements at the exchange are a forced reduction in leverage and hence in volatility.

My favorite trade in this situation would be a short straddle on medium term GVZ options, with a strike price of 20 or so. The volatility of GLD implied volatility is incredibly high, and unlikely to remain this high, so as vol-of-vol declines, profits are likely to accrue more quickly than usual to sellers; the downside strike incorporates the view that GVZ is just as likely to mean revert once GLD calms down as VIX does after an equity selloff. GVZ options aren’t active enough yet and the market is too wide, at least on the screen, but GVZ futures have been seeing steady activity and we’ve started a short (and admittedly early) position there. Selling cash-secured puts in GLD will also be an attractive trade once the metal finds a short-term bottom.

For reference, this is what implied vol in a currency normally looks like:

FXE 2 month and 1 year implied volatility, 2011-2013. Source: Livevol

There are periods when two month implied vol (yellow) is bid up to the level at which one year implied vol (off-white) is trading, but those periods are rare, and they don’t last long except in case of a real crisis. The September – November 2011 period stands out on this chart as an obvious instance of such a crisis. But otherwise, just as with equities, longer term vol demands a higher price.

Now compare the JPYUSD ETF (FXY):

FXY 2 month and 1 year implied volatility, 2011-2013. Source: Livevol

Until late December 2012, two month FXY options traded at a large discount to one year vols, i.e. the term structure of yen IV was in steep contango. But if you take the cross section of FXY IV in early January, you’ll find options at each tenor trading at roughly the same level, and that has continued throughout the first quarter of 2013, interrupted only by periods of backwardation.

First, some research on how great an idea it has been to sell one month straddles on equities since 2000:

Breakdown of P&L (index points) for systematically selling 1M straddles on E-STOXX50. Source: Deutsche Bank

Sell, sell, sell, and not just puts, either. The grey line there – selling 50 delta puts each month – has done better than SX5E over the period, but short calls really help a lot, especially during downturns. One of the findings of this study was that owning further out of the money short-dated puts to hedge your short straddle was not, at least on this index, obviously helpful in risk-adjusted terms. So the only reasonable conclusion would be to strap in and sell everything with a bid, right?

You can apply a naked short straddle view of the world to most assets. Take everyone’s favorite proxy for sectarian religious conflict these days, Japanese government bonds: disheveled Austrian prophets keep looking for doom and paying dearly for the privilege while the pudgy, well-coiffed parishoners of St. Maynard’s and St. Milton’s shuffle by. This is what normal looks like:

June 2013 Japanese government bond futures, March 6 – April 2, 2013, 30 minute bars. Source: TSE

A straddle jockey can make a mental note of that average true range (lower panel) around 0.03 – 0.05, or look at realized vols, or et cetera, and sell accordingly. It has been profitable. The research says so. Until:

June 2013 Japanese government bond futures, March 6 – April 5, 2013, 30 minute bars. Source: TSE

The doomsayers finally got a taste. Supposedly the April 5th action in JGBs was caused by a large Japanese bank or insurer that, surprised by the intensity of Kuroda’s plan, did not wait around to find buyers for all the bonds it wanted to sell. As a small bystander, when you’re into hour two of a selloff like that and buyers are still in hiding, there’s not much choice but to “manage risk”. The average true range now is three times your expected upper limit, your short call is helplessly without gamma, and your short put is eating up the last two months of profits and more. The only small grace is that, because you’re not short a whole strip of downside strikes (variance swap), your short put will run out of gamma, too, and eventually just act like a long futures position.

A clever trader replies here that the position ought to be delta-hedged. That’s fine – we do it – but five-minute jumps are hard to watch, especially when your research and your actions to this point assume at most a daily rebalancing. When people start taking bathroom breaks in shifts it is too tempting to change your “hedging regime” from daily to hourly to every crazy tick lower, which is not noticeably different, in P&L terms, from dumping the position or buying OTM gamma in the first place.

That’s enough, in my mind, to make perpetual naked straddles a non-starter, no matter what the backtest says. We’re not just talking about straddles here, either. Even though JGBs recovered in this case, even though the House finally passed TARP and Draghi said what needed to be said and the world will probably, hopefully, always have enough adults around to keep the bottom from falling out, the problem with strategies like this one is that they ask too much of us. No box is ever completely black and no strategy is ever completely automated as long as there is an off switch.

It doesn’t follow that the edge identified by even the most incurious option seller isn’t really there. The edge is as real as time and carry and the psychology of the average equity holder. The trick is in finding a sustainable way to exploit it, and while ideas are easy to come by, ad hoc practitioner intuition can never be as complete or as reliable as research. So we’ve come full circle: neither the nerds nor the jocks are self-sufficient, and success comes only in a synthesis.

Yet markets are sending Kuroda a clear message of their own. The cost of bearish options on Japanese shares is now at an eight-month high, according to Bloomberg News reports. In other words, investors who jumped into Japanese stocks in recent months doubt that the BOJ can deliver on its rhetoric. It should worry Kuroda that his vow to do “whatever it takes” to end deflation is already being questioned in financial circles.

The claim here is wrong in interesting ways. First, implied volatility does not entail a bearish outlook on asset prices. This is especially true at shorter timeframes. Presumably “the cost of bearish options” refers here to implied volatility. So, absent some additional claim about downside vs upside implied volatility skew, this is also a claim about the cost of bullish options on Japanese shares. So it does not follow that bearish investors are “sending Kuroda a clear message” via higher implied vol.

There’s also a factual issue here – Nikkei implied volatility is high, but not at an eight-month high. It’s lower now than the 28%+ levels seen in February. And as I’ve noted before, rapidly rising implied volatility over the last several months actually coincided with rising equity prices. It’s analogous to a VIX move from 12 to 18 alongside a 200 point SPX rally.

Nikkei 225 Volatility Index. Source: Nikkei, Condor Options

Additionally, positioning has to date been more about calls than puts:

NKY option open interest. Source: Bloomberg, BNP Paribas

The only bearish-for-Japan argument I can imagine based on option market activity would be that upside IV skew in JPY:USD CME futures has shifted over the last couple weeks. Higher one year call IV for 6J futures is evident here, but at SocGen noted recently, the most plausible explanation for this is that the most recent leg of the USD:JPY rally was remarkably less volatile (in historical standard deviation terms) than the moves we’ve seen since last October. That makes it less attractive to own unhedged dollar-yen upside, so the change in risk reversal/skew measures reflects a routine adjustment to recent price history, rather than a tectonic change in expectations. It’s also not the case that yen puts have gotten any cheaper.

Yen futures one year implied volatility skew. Source: CME, OptionWorks

USD:JPY 1Y risk reversal and spot price. Source: Societe Generale

The big bearish line has been, for months, that at some point all the jawboning in the world won’t be enough and the Japanese economy will have to start performing differently. So far, the evidence that such a moment has arrived  is pretty thin. It looks like Abe and Kuroda still have time. We are already long Japanese equities and short yen; investors who aren’t yet similarly positioned can take this opportunity to get in at better prices. The expectations game is always hard to play, but the risk/reward for the BoJ meeting still seems tilted toward the possibility that Kuroda may “over-deliver” and surprise markets. If Kuroda does disappoint, the next-best approach will be to harvest that still-rich equity and dollar/yen upside skew.

Source: “Is Haruhiko Kuroda Japan’s Mario Draghi?” Bloomberg, April 1, 2013