Money Math

What The Widowmaker Can Teach Us About Trade Prospecting And Fool’s Gold

We’re going to go on a little ride to talk about trade prospecting. We’ll use the natural gas futures and options market to demonstrate how to think about markets and what’s required to actually identify opportunities.

The nat gas market is all the rage these days as we head into the winter of 2021/22.

Let’s start with some background.

The Widowmaker

Enter the famous March/April futures spread in the natural gas market. This was the football famously tossed between John Arnold’s Centaurus and Brian Hunter’s Amaranth. You can get a good recount of the story here as recounted by the excellent @HideNotSlide.

The reason it’s a “widowmaker” is the spread can get nasty. The March future, henceforth known by its future code (H), represents the price of gas by the end of winter when supply has been withdrawn from storage. April (J) is the price of gas in the much milder “shoulder” month. H futures expire in Feb but are called “March” because they are named by when the gas must be delivered. Same with J. They expire in March, but delivered in April. The H/J spread references the spread or difference between the 2 prices.

If you “buy” the spread, you are buying H and selling J.

If the price of the spread is positive, the market is backwardated. H is trading premium to J.
If the spread is negative, H<J (ie contango)

On 10/6/2021 the spread settled at +$1.44 because:

H future = $5.437
J future = $3.997

Introducing Options Into The Mix

There are vanilla options that trade on each month.

So there are options that reference the March future and they expire a day before the future (so in February).

H settled $5.437 so the ATM straddle would be approximately the $5.45 strike. Strikes in nat gas are a nickel apart.
For April futures the ATM strike is the $4.00 line. You can see the J straddle (ATM C + P) settled around $1.14

Commodities Are Not Like Equities

Every option expiry in equities references the same underlying — the common stock price. If you trade Sep, Oct, Nov, or Dec SPY options they all reference the same underlying price.

The December 100 call cannot be worth less than the November 100 call because of simple arbitrage conditions. Your December options also capture the volatility that occurs in November (in fact if you wanted to bet on the volatility just in December, you would structure a time spread that bought December vol and sold November vol, to strip out all the time before November expiration. The structure of that trade is beyond the scope of this post.)

This doesn’t work in commodities because each month has a different underlyer.

Recall H =$5.437 and J = $3.997

The H $5 call is almost .44 ITM
The J $5 call is a full dollar OTM

Despite J options having a month longer until expiry, the J $5 call trades waaaay under the H $5 call.

It gets better.

Even if H and J were trading the same price, the H $5 call can trade over the J $5 call. This is where newcomers to commodities from equities find their muscle memory misfires.

The H implied volatility can go so far north of the J vol that it can swamp the 1 month time difference.

As described earlier, in an equity, March and April options would reference the same underlyer so owning April vol exposes you to the March vol.

Not true in NG.

Severing the arbitrage link between spreads

Backwardation

H is trading above J. The spread is backwardated. But H and J are not fungible. They are deliverable at different times. If you need H gas, you need H gas. It’s cold today. You cannot wait for J gas to be delivered. You won’t need it then.

This is generally true in commodities.

There is no arb to a backwardated market.

Contango

A contango market can be bounded by the cost of storage. Be careful though. The steep contangos of oil in Spring 2020 and around the GFC are lessons in “limits to arbitrage”. The cost of storage is effectively infinite if you run out of storage. So contango represents the market “bidding for storage”. You can’t just build new storage overnight. The other major input into contango spreads is the funding cost of holding a commodity either via opportunity cost or interest rates. THE GFC was a credit crunch. Funding was squeezed. That cuts right to the heart of “cost of carry” that contango represents.

So we now understand that H and J can become unhinged from each other. That’s why the spread is a widowmaker. It can be pushed around until convergence happens near the expiry of the near month. That’s when reality’s vote gets counted.

More Complexity: Options On Those Crazy Spreads

You can also trade options directly on the H/J futures spread. Since H/J is considered a calendar spread, the options are cleverly named:

Calendar spread options.

The cool kids refer to them as “CSOs”.

Let’s talk CSOs.

We established that the H/J future spread is $1.44

You can buy a call option on that spread. You can buy (or sell) an OTM call, like the H/J $10 call.
You can buy an ITM call like the H/J $1 call. That option is 44 cents ITM.
You can buy a put on the spread. If you buy the H/J 0 put (pronounced “zero put”), that option is currently OTM. It goes ITM if H collapses relative to J and the spread goes negative (ie contango).

These exist in WTI oil as well. Imagine a fairly typical market regime where oil is in contango. The CL1-CL2 spread might trade -.40. That means the front month is .40 under the second month. CSOs trade on these negative spreads as well! If someone buys the -$1.00 put they are betting the market gets even more steeply contango.

I’ll pause for a moment.

Right now, you playing with an example in your mind. Something like: “so if I buy the -$.25 call, I’m rooting for…ahh, CL1 to narrow against CL2 or even trade premium into backwardation”

Don’t be hard on yourself. This is supposed to hurt. It hurts everyone’s head when they learn it. It’s just a language. The more you do it, the easier it gets and with enough reps you won’t remember what it was like to not be able to understand it natively.

Real-life example

These prices are from 10/6/2021 settlement.

H settled $5.437

The H 15 strike call settled $.42

H/J spread = $1.44

H/J $10 CSO call = $.38

Let’s play market maker.

You make some markets around these values:

Suppose you get lifted on the CSO call at $.40 (2 cents of edge or 20 ticks. 1/10 cent is min tick size)
Meanwhile the other mm on your desk gets her bid hit on the vanilla H 15 call at $.40 (also 2 cents of edge)

Your desk has legged getting long the H 15 call, and short the H/J 10 call for net zero premium. If we zoomed ahead to expiration what are some p/l scenarios?

H expires at $5 and J is trading $4 on the day H expires or “rolls off”. Therefore H/J = $1
- Both calls expire worthless. P/L = 0
H expires $15 and J is trading $4 so H/J is $11.
- Ouch. Your long call expired worthless and your short H/J $10 call expired at $1.00. You just lost a full $1.00 or 1,000 ticks. That’s a pretty wild scenario. H went from $5.43 to $15 and J…didn’t even move?!

How about another scenario.

H goes to $16 and J to $7. So H/J expires at $9.
- The $10 CSO call you are short expires OTM and the vanilla H 15 call earned you $1.00. Now you made 1000 ticks.

It quickly becomes clear that vol surfaces for these products are untamed. Option models assume bell-curvish type distributions. They are not well-suited for this task. You really have to reason about these like a puzzle in price space. I won’t really dive into how to manage a book like this because it’s very far out of scope for a post but it’s critical to remember that pricing is just one consideration. Mark-to-market, path, margin play a huge role.

Sucker Bets

The truth is the gas market is very smart. The options are priced in such a way that the path is highly respected. The OTM calls are jacked, because if we see H gas trade $10, the straddle will go nuclear.

Why? Because it has to balance 2 opposing forces.

It’s not clear how high the price can go in a true squeeze or shortage
The MOST likely scenario is the price collapses back to $3 or $4.

Let me repeat how gnarly this is.

The price has an unbounded upside, but it will most likely end up in the $3-$4 range.

Try to think of a strategy to trade that.

Good luck.

Wanna trade verticals? You will find they all point right back to the $3 to $4 range.
Upside butterflies which are the spread of call spreads (that’s not a typo…that’s what a fly is…a spread of spreads. Prove it to yourself with a pencil and paper) are zeros.

The market places very little probability density at high prices but this is very jarring to people who see the jacked call premiums.

That’s not an opportunity. It’s a sucker bet.

Let me show you what’s going on with the CSOs:

The CSO options tell us that the H/J spread has roughly 3% chance of settling near $2, a 2% chance of ending near $3 and a 0% chance of settling anywhere higher than that.

And yet the futures spread is trading $1.44 today! And the options fully expect that to collapse.

What is going on?

Look at history. Even in cold winters, the spread almost always settles….at zero! When H expires, it is basically going to be at the same price as J.

Now, I know nothing of gas fundamentals. And none of this is advice. And I’m not currently up on the market, but I am explaining how these prices look so crazy (as in whoa look at all this opportunity) but it’s actually fair.

The market does something brilliant.

It appreciates path while never giving you great odds on making money on the terminal value of the options.

The Wider Lesson

So how do you make money without a differentiated view on fundamentals in such a market?

There are 2 ways and they double as general lessons.

Play bookie
You have a team that trades flow. You are trading the screens and voice, you’re getting hit on March calls over here, you’re getting lifted on March puts over there, you’re buying CSO puts on that phone, your clerk is hedging futures spreads on the screens. Unfortunately, this is not really a trade. This is a business. It needs software, expertise, relationships. Sorry not widely helpful.
Radiate outwards
The other way to make money is prospecting elsewhere, with the knowledge that the gas market is smart. It’s the fair market. It’s not the market where you get the edge, it’s the one that tells you what’s fair or expected. So you prospect for other markets or assets that have moved in response to what happened in the gas market, but did so in a naive way. A way that doesn’t appreciate how much reversion the gas market has priced in. Can you find another asset that’s related, but whose participants are using standard assumptions or surfaces? Use the fair market’s intelligence to inform trades in a dumber or less liquid or stale market.

Trading As a Concept

Many people think that trading is about having a view. Trading is really about measuring the odds of certain outcomes based on market prices. Markets imply or try to tell us something about the future. The job is to find markets that say something contrary about the future and take both bets. Arbitrage is an extreme example of this. If one person thinks the USA basketball is 90% to win the gold and another thinks the field is 15% to win the gold you can bet against them both and get paid $105 while knowing you’ll only owe $100. Trading identifying similar examples but of course in reality they are hard to find, more difficult, and require creativity and proper access.To see the present clearly you must be agnostic. You look for contrary propositions. Trading is not about having strong opinions. It’s not thematic. You don’t have some grand view of what the future looks like or the implications of some emerging technology or change in regulations. You just want to find prices that disagree.

Why would you slug it out in smart markets? Use them to find trades in markets that radiate away from them that are not incorporating parameters from the smart market fully. If you can’t get away from fair markets, you are going to need to be absolutely elite.

Battling it out in SPY reminds me of this cartoon:

The solutions in markets are rarely going to be where it’s easy to see because that’s where everyone will be looking.

Happy prospecting.

If you found CSOs interesting recognize there are physical assets that are just like options on a spread.

Oil refineries =Heat/Gas crack options
Power plants = Spark spread options
Oil storage facility = WTI CSO puts
Soybean mill that crushes soy into meal/bean oil

If you had a cap ex program to build one of these assets how would you value it? You’d need to model volatility for the spread between its inputs and outputs!

The owners of these assets understand this. They are the ones selling CSOs! It’s the closest hedge to their business.

I got the data for this post from the CME website’s nat gas settlements page.

The dropdowns on the right of the page should keep you busy.

If you use options to hedge or invest, check out the moontower.ai option trading analytics platform

editor