Quantifying the Brexit premium and the Walking Dead, Season 8

Before addressing today's FOMC meeting Macro Man wanted to briefly return to the Brexit issue because it really is so very interesting.  One thing that he can not stress enough is the ephemeral nature of market pricing.  Yesterday's post cited 2 week vols of 26%, which is what Macro Man pulled from his Reuters feed.   Yet Tuesday morning they were getting paid at 42!

According to one of Macro Man's mates in the market (who occasionally comments here), yesterday morning the implied overnight vol for the 23rd/24th was 140%, which translates to roughly 6%, or 830 pips.   Yet by the end of the day another reader was claiming that the overnight was now priced at 8%, which represents an implied vol level of 190%.   Needless to say, these types of volatilities are essentially unprecedented in foreign exchange, perhaps excepting long-distant crisis events such as USD/IDR in 1997 or DEM/ITL in 1992 or 1995.

In any event, sterling vols are now firmly in Rolls Royce territory: if you have to ask the price, you probably cannot afford it.  Macro Man thought it might be interesting to try and tease out some data on the market impact of swings in Brexit expectations.   His first port of call was to use some sort of control model; he looked at the 2y2y spread between US and UK rates (chosen to fall largely outside of a putative Brexit) as an explanatory variable for cable, using 2014-15 as an in sample period.  Gratifyingly, he found a generally good fit, but with cable trading substantially lower than the model would suggest over the past nine months.   There is our Brexit premium!


(Note that the usual caveats apply: the regression was run on levels rather than changes given the noisiness of high frequency data, etc.)

His mate then sent him the Brexit betting odds helpfully stored on the Bloomberg terminal.  Running the model for this year obviously generates a much better fit; importantly, though, recent changes in the model have echoed the move in spot the way that a simple rates model does not capture.  Interestingly, the new model is now at its most bearish on cable- unsurprising, really, given where the betting odds are!

The model stats are set out below.  As you can see, in 2014-15 a 100 bp relative shift in 2y2y swap rates in the UK and US was worth more than 21 big figures; this year, it's worth slightly less than 10.   The real interesting bit is the beta of the Brexit odds: 0.264.  The implication, therefore, is that given how the market has traded off of betting odds, the ultimate difference in cable based on whether Remain or Leave win is 26.4 big figures.   Given the current odds of 43.3%, the upshot is that rate spreads being equal, cable will sell of 15 big figures on a Brexit vote (call it 1.25 or so) and rally over 11 figures on a Remain vote (1.52-ish).  In a world where we don't know if the implied overnight is worth 140 or 190 or 220 vols, these numbers really don't seem too far out of whack.  With the sensitivity of cable to swings in polls and expectation, it sure feels like there's some value in owning gamma that expires even before the vote, given the illiquidity in spot and gappy price action.   Good luck finding a friendly vendor at a reasonable price, however....

As for the Fed, yesterday's fixed income price action felt more than a little climactic, helped partially by Bunds having a decent go at rejecting negative yields.   The candle on the 10y yield chart certainly looks like a reversal pattern, which would jive nicely with the Fed grasping the wheel of the monetary bus and speaking calmly and confidently over the intercom.



While there's always a source of worry (how can there not be, when one of your self-appointed decision criteria is "global conditions"?), despite the recent payroll number the economy is tracking along very nicely in Q2, thank you very much.  Yesterday's retail sales figure was solid, with the retail control figure positing its highest 3m/3m change in nearly two years.


Of course, over the last few years it's taken five pieces of good news to counterbalance one piece of bad, and the payroll figure was clearly pretty bad.   That being said, recent Fed communication has done little to suggest anything but a general (though not universal) desire to adjust rates upwards before too long.

Now, it's clear to everyone that the only two hopes of a rate hike today are Bob Hope and No Hope, so clearly attention will be paid to the forward looking message from the statement, the SEP, and the press conference.   Given the extremes in market pricing (driven both by Brexit fears and the black hole of yield located in Frankfurt), now might be an apt time to gently lean against market pricing (or at least, expectation) so as to keep the committee's options open.

Macro Man was amused and bemused to see another piece from Larry Summers trotting out the usual OH NOES, NOT RATE HIKES! trope.  Macro Man had to chuckle at the threat that in hiking rates, the Fed could push the US into the same boat as the EU and Japan, both of whom have taken rates negative and seen their banking sectors eviscerated as a result.

There remains no empirical example of a large heterogeneous economy escaping extreme disinflation through zombification, crushing the cost of (and return on) capital, and encouraging investment into low-return pursuits that hamper productivity.  There remain numerous examples of these policies failing to achieve the inflaton goal.  The Walking Dead gets great ratings from academics but it sucks to live within the narrative.

Macro Man feels compelled to trot out one of his favourite charts which summarizes the current situation to a T.  Simply put, the data would suggest that the dearth of inflation in the US can be explained simply by a prior rapid expansion in the country's industrial capacity, with the level of that capacity well above trend and utilization rates well below historic norms.  Can we really expect a secular uptick in inflation with so much excess capacity?


All of this will get you a cup of coffee and little else, however.   What matters is what the Fed will say and do.  For choice, Macro Man does expect them to acknowledge the risks posed by Brexit and elsewhere as a primary short-term focus.  Ultimately, however, he looks for them to say with that caveat, if things go smoothly and the economy performs as expected (ie, continues tracking 2.5% + and the labor data snaps back) then the door remains wide open for another rate move before long.

Whether markets like that may well be irrelevant in the short term, with option expiry, the looming vote, and a scramble for vol and safety likely to dominate in the coming days.
Previous
Next Post »

33 comments

Click here for comments
JB
admin
June 15, 2016 at 8:16 AM ×

Has anyone seen Sky news this morning with the England football......... Sry, I couldn't help myself.

V interesting piece MM.

There has been some aggressive selling into the European cash closes the last few sessions, hasn't quite felt like a classic capitulation low, but then again the entire sell off in Jan/Feb was all very orderly intra day. In the short term though Dead cats do bounce.

Reply
avatar
Anonymous
admin
June 15, 2016 at 8:51 AM ×

I mentioned 8% overnight move, which is 130 vols (130 divided by square root of 252 is a bit more than 8%), not 190, so it's consistent with what your other friend says. It doesn't really change the point, but it's slightly less volatile than portrayed.

Reply
avatar
Thud and Blunder
admin
June 15, 2016 at 9:13 AM ×

Interesting read; thanks, MM.

On US industrial capacity, I have a had time swallowing that the US expanded their industrial capacity by circa 50% between 1994 and 2000, even with tech booming. Even looking at subsectors, you see a similar pattern in manufacturing. This doesn't seem to pass the smell test, nor does my admittedly hazy recollection of the 90s recall a US manufacturing renaissance. Are you comfortable placing weight on this data?

Reply
avatar
Macro Man
admin
June 15, 2016 at 11:51 AM ×

@anon 8.51, your math is wrong. An overnight vol of 24 corresponds to 1% in the straddle, and a vol of 140 corresponds to 5.83% in the straddle...or 830 pips.

Reply
avatar
Anonymous
admin
June 15, 2016 at 11:52 AM ×

The mdoel is a bit too simple, it implies that cable would be trading significantly different when odds for Brexit were at let's say 70% or 90%, I don't think that would be the case. I think a Brexit would be more or less fully priced in at 60% Brexit vote probability. The most violent reaction to polls is in the 40-60% Brexit probability area. I don't have a mdoel but think GBP/USD would settle around 1.33 in case of Brexit

Reply
avatar
Booger
admin
June 15, 2016 at 1:04 PM ×

Cable 1.35 area is the area of interest if Brexit fears ripen. I think that was the GFC low.

Bonds look more appetizing today and I do wonder whether yesterday was a reversal candle. That has me nibbling on a short position for longer dated stuff today. Retail sales in line, Fed likely to say something along the lines of carry on. Brexit risk all out there. 30 year may have topped out in the short term.

Captcha test on click all the store fronts has me failing on multiple occasions. My gotten, is it getting harder ?

Reply
avatar
washedup
admin
June 15, 2016 at 1:24 PM ×

@MM - that larry summers article (frankly his views in general) you linked just made me angry - how can bright people be this stupid - they won't accept the possibility their growth models for the last 40 years may be wrong, so lets just raise the inflation target, not raise rates, and do some infrastructure spending and voila, japan no more. Just f@#ng accept sometimes you don't get growth and live with it - but alas, too many promises have been made to the contrary by his political brethren to avoid pitchforks, so growth at all costs it is.
As for the inflation targeting, its as if the 360 LB man has a better chance of losing weight merely by lowering his ideal weight target to 200 from 250 without any change to lifestyle.
@Boog re long bond - when the RSI of something goes to 90 you do tend to get pullbacks - to me it seems just as likely it may consolidate before going on to print all time highs, so Im staying out of it - I will say your opinion on it is consistent with your thoughts on crude.

Reply
avatar
Booger
admin
June 15, 2016 at 2:25 PM ×

Washed: good point, wish I'd thought that before I entered as I'm starting to regret it already! It does feel like everyone is thinking bonds can only go one way though and positioning is getting one sided with Brexiters, ecb piggybackers getting onside with the usual bond bulls. Perhaps some confirmation would have been more prudent and I may well be proved to have been a week or 2 early (and wrong) there.

Aargh...the storefront reCaptcha test again.

Reply
avatar
Anonymous
admin
June 15, 2016 at 4:15 PM ×

MM, unless by overnight you mean a smaller timeframe than a daily move, then 16 vols, not 24, equals 1% daily move. Regardless of semantics, the quoted vols (130) are for the daily move (23/24 1 day forward vol), so daily move of 8%.

Reply
avatar
Macro Man
admin
June 15, 2016 at 4:34 PM ×

No, you're wrong. For an option priced at time t, with an expiry of t+1, a vol of 24 corresponds to a 1% straddle price. Any option price will verify this. I am not sure why you used the sqrt of 252 above, as options still have worth (and you still pay decay) on non business days. BTW, a 1y straddle at 130 vol is not worth 130%, a 1y at 24 is not worth 24%, etc etc.

Reply
avatar
Macro Man
admin
June 15, 2016 at 4:35 PM ×

Any option pricer, that should say. The 24 vol shorthand was one of the first things I learned on the CME currency option floor when I entered the workforce.

Reply
avatar
washedup
admin
June 15, 2016 at 4:43 PM ×

@anon - MM is correct - the straddle price is P*sqrt(T)*0.8*vol, with P being the underlying - since sort(365) is roughly the same as 24*0.8 - the above reduces to P when vol is 24% - hence the identity.

Reply
avatar
Anonymous
admin
June 15, 2016 at 4:44 PM ×

Options are quoted in price not in vol. Vol is model dependent, in my experience we always used holiday calendars (including weekends). Hence for me ~16 vol is roughly 1% move.

Reply
avatar
Anonymous
admin
June 15, 2016 at 5:02 PM ×

Monetary policy would be more effective in a world where prices increase at 4% per year rather than 2%.

Reply
avatar
Nico G
admin
June 15, 2016 at 5:05 PM ×

the anons are right

you use 252 days for hedging purposes - you can only hedge your delta when market trades

so if you loaded gamma at 16 vol you will need 1% move close to close (most players hedge at close) to break even hence the 16/1 rule used by gamma players and market makers

on that note i covered the remaining 200/370 of Estoxx sep futures this morning made a million or so and am taking the summer off

i wish you all good luck and be very careful around Brexit date

Reply
avatar
abee crombie
admin
June 15, 2016 at 5:09 PM ×

Thanks for the nice post MM.

Nico you are my hero !

Reply
avatar
johno
admin
June 15, 2016 at 7:28 PM ×

6 dots for 1 hike this year. That's likely where the real voting consensus is. Might as well be zero hikes ...

Reply
avatar
Bruce in Tennessee
admin
June 15, 2016 at 7:29 PM ×

Interesting fed response..the fed announces no rate hike and a downgrade of the economic outlook and the DOW drifts lower...putting that together, perhaps traders are thinking that NIRP is coming here as the economy sputters (?), and that banks, pension funds, non-risk takers, you know the list...are going to do less well in 2016-2017. Certainly the lack of a reflex bounce is interesting....


The Mirkwood economy?

Reply
avatar
Anonymous
admin
June 15, 2016 at 7:34 PM ×

I'm shocked ! The Fed are gonna leave rates on hold !!!!

Hahaha did any stupid fcks seriously think these retards would ever raise rates again?

Reply
avatar
Macro Man
admin
June 15, 2016 at 7:54 PM ×

@ Nico, no they are wrong, and so are you. If you ring up your friendly FX dealer and sell overnight vol at 16, you will be very very sadly disappointed when you find that you receive 66 bps rather than 100 bps for the straddle.

The move you need to hedge is not what is under discussion...what we are talking about is the price that you trade at, or what the TV of an overnight straddle of a particular vol is. The TV doesn't give a shit whether you can hedge or not.

Reply
avatar
abee crombie
admin
June 15, 2016 at 8:03 PM ×

"um we dont really know where rates should be"


Reply
avatar
Polemic
admin
June 15, 2016 at 8:11 PM ×

The no mans land between Theory and Reality is littered with the bodies of quants.

Reply
avatar
Macro Man
admin
June 15, 2016 at 8:16 PM ×

Well POl, as you well know, if you ring up an FX dealer and give 16 in the straddle and demand 1% premium because the equity boys have a rule of thumb, you'll get a two word response, the second of which is "off".

Reply
avatar
Polemic
admin
June 15, 2016 at 8:21 PM ×

Absolutely MM. The price is the price you can trade on not what your computer thinks it is. And we haven't even added the lower bid than expected function that quants can't work out but we call the "He's already called Mickey at xyzbank and passed a low one" factor.

Reply
avatar
Anonymous
admin
June 15, 2016 at 8:23 PM ×

So ignore the fed Labor Model... just focus in initial claims and "perceptions" of the labor market. She is kidding, right?

Asked about why she left out rate hikes in the coming months in speech:
"We need to make sure there is sufficient momentum"

How much more can we tolerate from Yellen...
.

Reply
avatar
Macro Man
admin
June 15, 2016 at 8:27 PM ×

She more or less admitted, "we don't know what's going on or what to do, so let's bury our heads in the sand and hope for the best."

Reply
avatar
Anonymous
admin
June 15, 2016 at 8:33 PM ×

As MM (and others) have said before, the Fed under Yellen has zero credibility.

Reply
avatar
Anonymous
admin
June 15, 2016 at 8:37 PM ×

If there are any "Traders" left out there, the "Market" will call the FED out on this sh!t.

Reply
avatar
washedup
admin
June 15, 2016 at 9:16 PM ×

@anon 8:37 will the "traders" do this by moving the "markets" using a sophisticated light beam called the "LASER"???

Now on to the BoJ, which was last seen lurking around the event horizon of a black hole - anyone expect anything profound out of that one? Funny how JPY vol at 32 back in Feb actually turned out to be under, not overpriced into the NIRPageddon.

Reply
avatar
Polemic
admin
June 15, 2016 at 9:48 PM ×

Washed.. that made me giggle. The Austin Powers of the dealing room.

Reply
avatar
Leftback
admin
June 16, 2016 at 1:53 AM ×

One of the stranger Fed days of recent years. La Paloma Blanca was certainly more clueless than usual today, but nevertheless relentlessly dovish, and yet Mr Market decided to sell.

OpEx likely to be the next event to nudge the markets up and down, we'll see how bold the vol sellers are going into the weekend. The dollar's reaction today was also interesting. La Paloma Blanca clearly was priced in by FX punters, and there were even a few dip buyers in the greenback today. We note that oil now goes down every day.

Can't help wondering if the "Shanghai accord" earlier this year largely consisted of Yellen telling China to quit trying to devalue CNY, sell the fricken USD and buy cheap crude and store it on ships, thereby relieving the pressure on the US and reflating the global economy for a while at the same time.

Sitting it out for a few days in equities, but short crude and long UUP.

Reply
avatar
Unknown
admin
June 16, 2016 at 10:27 AM ×

The vol quoting convention is interesting. In the equity world we always use a factor of sqrt(252) = 16ish to annualise, as the vol is quoted on a 5/7 basis: working days only. (The same vol, quoted on a 7/7 day basis, is obviously lower.)

What's the reasoning behind the x24 calculation when annualising fx vol?

Reply
avatar
Anon
admin
June 16, 2016 at 12:19 PM ×

I think I understand the x24 convention -- it interpolates between 365-day annualised volatility and the daily break even:

Recall that the atm approximation of BS is 0.4 sigma spot sqrt-time (pricing a call and, equivalently since we are at the money, also a put). An atm straddle therefore costs twice as much = 0.8 sigma spot sqrt-time, expressed in dollars. Expressed as a percentage of spot, the price is 0.8 sigma sqrt-time. Of course, the price of the straddle is exactly its breakeven. Suppose a 1year atm straddle costs 10%; the formula tells you the volatility realised over 1 year is expected to be 12.5%. A one-day atm straddle costing 1% has an implied daily vol of 1.25%. Annualise 1.25% (over 365 days) is equal to 1.25%x19 = 24%.

Hence 1% daily breakeven is equivalent to 24% daily vol. I think forex traders usually quote yearly vol and daily breakevens, although these are slightly different concepts. Equity traders quote yearly vol and daily vol. Hence we have different formulas to convert between daily and annual: different things are being converted.

Reply
avatar