Authored by Kevin Muir via The Macro Tourist
Today I have decided to try something a little different, but before we start with my experiment, let me sketch out what’s driving this line of thinking.
I have long believed the next financial crisis will not emanate from risk assets, but instead will result from governments losing control of their bond markets. I am a big fan of Bill Fleckenstein’s line:
Sure, we will have risk hiccups as Central Banks attempt to normalize rates and monetary policy, but the big one - you know, the one no one is prepared for - will occur when easing causes the long end of the yield curve to sell off hard. At that point, there will be no easy answers for politicians and policy makers. That’s the end game.
Therefore, although I am sympathetic to the idea that risk assets will be volatile, I think the better volatility-buy is supposedly riskless assets. Yup - I think owning sovereign debt volatility is a better way to capitalize on the coming financial instability.
Bond vol is dirt cheap
Great. The ‘tourist has a view about the next financial crisis. So what? Opinions are like elbows - almost everyone has one, and most of us have two.
How does this translate into an actionable idea?
Let’s first have a look at U.S. treasury bond volatility and see if everyone else has already discounted this view.
Well, cover me in jam and tie me to an anthill - far from the market agreeing with me, it looks like treasury bond volatility hasn’t been cheaper in decades.
Why is that?
There are a variety of reasons. They include various global Central Banks pegging different parts of the curve or engaging in monstrous quantitative easing programs. It’s also a function of the Federal Reserve’s shift of the past couple of decades towards more transparency and clear forward guidance. And finally, the Fed took out a fair amount of volatility from the market place when they bought their huge slug of mortgage-backed-securities and didn’t delta hedge them.
Yet I contend the longer an asset is pegged, or manipulated, the larger the imbalance and ultimately, the larger the break. And nothing has been suppressed more than the price of money.
Fixed income managers are notorious vol sellers
However, there is another dangerous aspect to this setup. Years of bond stability has created an environment where bond managers have embraced selling volatility to earn extra return. There is one famous bond manager who when he isn’t filling his ex-wife’s house with dead fish smells (allegedly), loves to sell bond volatility. It seems like every time I see him interviewed, he is giving his view, but then adding how he is juicing returns by selling options on outcomes that he doesn’t think will occur (see “Bought from you Bill” and “Bund bears in their natural habitat”). And it’s not just Bill. Don’t forget the whale leaning heavily on 3-month Eurodollar volatily (”The Monster Eurodollar Option Trade”). Selling volatility is ingrained into the psyche of bond managers.
Don’t misunderstand my argument. I get it.
I completely understand why they are selling volatility. I cut my eye teeth on an institutional equity derivative desk. Equity index implied volatility typically trades fat relative to underlying volatility, so when you work for a bank with a huge balance sheet, you are often short vol. It makes sense. You have the best ability to delta hedge. You have the most efficient execution. You have a balance sheet that (theoretically) should be able to withstand adverse movements.
So I “grew up” short vol. And I can tell you honestly, I am way more comfortable earning the theta rather than paying it out. You don’t know pain until you have a huge monster slug of long theta that is eating your P&L every slow summer day (I guess you might argue that being short gamma in a nutbar market dislocation is more painful, but at least shit is happening and you aren’t losing money for no reason).
Anyways, I don’t discount why fixed income managers have been such large sellers of vol during the past decade. After all, it has worked.
But I will recount one of my favourite Warren Buffett quotes (which I probably overuse):
We have hit the point where shorting bond vol is a dangerous game.
But how do you buy bond vol?
Which brings me to today’s experiment. There is no easy single instrument (that I know of) which allow you to buy fixed-income volatility like the VXX ETF does for equity index volatility.
As a quick aside, buying VIX products - whether they be ETFs or the VIX futures is not truly betting on increased volatility. It is a bet on the market’s pricing of future implied volatility. I know that seems like a minor distinction, but it is important to understand that VIX products are based on market’s expectations of forward volatility as opposed to actual underlying volatility. Now obviously future volatility is often best forecasted by recent past vol, so there is a relationship, but it doesn’t necessarily go together one for one.
Back to our bond vol problem. With no easily tradeable instrument, we are stuck with a problem about how to get long.
It’s not nearly as complicated as many investors envision, but I understand there is some option pricing theory involved, so I have decided to walk you through the process, and most importantly, update it every day for the life of the position. So this post will be refreshed every day as the market conditions change and will hopefully give you a better understanding on how option trading works.
December 118 straddle
Where do we start?
Well, late last week, my favourite futures broker pal, Alex Manzara noted the following:
TYZ 118 straddle traded 1’25 or higher on last week’s turmoil, but settled below 1’00 at 0’63 with 37 days to go. This period covers the employment data, US midterm election, and Nov 8 FOMC.
Alex - I couldn’t agree more. We have a ton of catalysts coming due in the next month, yet bond vol is staring at us like a sassy teenager smacking her gum while asking us, “So big shot, whatdaya wanna do?”
I want to buy vol. Big time.
So, let’s do it. And if this is going to be paper-trading anyway, let’s trade like Soros or Druckenmiller and buy a yard of 118 straddles.
I am going to initiate the trade at last night’s closing prices.
The TYZ8 118 Call went out ‘31 bid at ‘32 ask (option trades are quoted in 64ths). The 118 Put was ‘33 at ‘34. The straddle was therefore ‘64 at ‘66.
Now we could try to bid in the middle, but buying 10,000 probably takes paying the offer (even in my dreams I need to be somewhat realistic), so let’s take a fill at ‘66.
Therefore we own 10,000 TY Dec 118 Calls at ‘32 and 10,000 TY Dec 118 Puts at ‘34.
What’s the next step? We could just pull a Richard-Dreyfus and let it ride. If the futures closed more than 1 and 1/32nd away from 118, then it would be profit. Anything closer to 118 would be a loss.
But that’s not what we are trying to accomplish. We want to be long vol and delta hedge it so that regardless of the ending value of the underlying future, if the historical volatility during that period is more than the implied volatility that we paid, we win.
Think about it. What if during the next month the 10-year future plunged to 115, then rallied like a banshee up to 123, only to close right at 118? The long vol trader who didn’t delta hedge would be out their entire premium, yet the delta hedger would be rubbing their hands in glee.
Our choice is pretty clear - we need to hedge, but how do we go about that?
There are a bunch of different ways to accomplish hedging and this is where the art of option trading comes into play. Some like to hedge when their position reaches a certain delta. Others like to hedge at a certain time of day. Personally I would never hedge at the same time each day, and let me tell you why. Way back when I was on the trading desk, there was another dealer with a big short option position. In Canada, brokers all have numbers that they disclose down on the exchange to signify which firm executed the trade. This particular broker had a number that corresponded to a famous hockey player - Tim Horton (yup, for all you non-Canadians, our national coffee chain is named after a hockey player - what isn’t named after a hockey player in Canada after all?) Anyways, this broker would execute every day at 330pm. The whole desk would whoop it up as we waited for Tim Horton to come blasting out his hedges at 330pm everyday. We would cheer as Tim never disappointed. Eventually I figured out his position (I calculated his gamma based on his trades) and then I would pre-buy or sell (based on the change in the index on the day) his position at 325pm and stuff him at 330pm.
So I would never execute our delta hedge at the same time each time, but this is all pretend, so let’s make our life easy and hedge it based on the futures close at 3pm EDT each day.
Now comes the next little option trading nuance.
We are getting long volatility because we think the market is underpricing the future volatility of the underlying. Therefore, why would we use the current implied volatility estimate to calculate our hedges? I don’t get this rookie mistake too many option traders make. Don’t let your greeks be implied from the market prices. If you do that, you will be allowing market vagaries to affect your hedging.
Our option straddle is trading at approximately 3.7% implied volatility, but I think it is worth at least 4%, so that’s what I am sticking in my model.
How did I come up with that?
It’s a fair amount of guessing, but this chart might help.
You can see that volatility has rarely been this low, and 4% is a chip-shot. Especially with all these catalysts on the horizon.
Plugging the 4% vol into the option calculator comes up with a delta of 49.225 for the calls and -50.5654 for the puts. Therefore on our 10,000 lots, we end up having a delta of
10,000 x 49.225 = 4,923 long
10,000 x -50.5653 = 5,056 short
To delta hedge, we need to buy an extra 142 contracts. We will do that at 117’31.
There we go. That’s the position. I am going to monitor it and put in the delta hedges at every close. I will also calculate the P&L to walk you through how it works.
Please come back and join me on my experiment of maintaining a large long bond vol position in real time. Updates will occur between 4 and 5pm EDT.
Thanks for reading,