European energy and political risks: Is it too early to talk about 2027?

CPI : Triple Z

Absolutely nothing to see here on CPI and now we enter the last few weeks of summer with nothing particularly delicious to chew on. September FOMC is still 50/50 and presumably comes down to the jobs report and CPI reports released September 4 and September 11 respectively. Core PCE is too easy to predict at this point for that release to matter.

And while you could, in theory, get excited about Jackson Hole, my guess is that this year’s meeting in Wyoming will be a nothingburger as Warsh’s words have lost meaning and the headline for the show is: Financial Innovation: Implications for Payments and Policy. That’s a setup for a wonky wonkfest.

If nothing is going to happen between now and Labor Day, is there a way to make money in a market where FX vol is already in the basement? You can do a carry trade for a few weeks. You are not going to get rich doing so, but you will at least pay your technology bills. For example, if you sell 100 EURHUF here and nothing happens, in a month you have made $275,000 of profit.

Much less scary: If you buy 100 USDCNH here and nothing happens, in a month you have made $200,000. These are just examples of the types of things people are doing right now because there is not much going on and vol is too low to sell. I don’t mind these sorts of trades as we could easily have another 18 calendar days (or more) of ZZZ. Labor Day is late this year, too, so the normal pickup in interest and engagement for the sprint to the year end finish comes later than usual. Labor Day is September 7 this year, the latest possible date for that holiday.

Two things I am thinking about

With nothing really happening in markets today, two themes on the horizon that are worth thinking about:

1. Brazil elections. The runup into the Brazilian election could make for a few interesting trades. I will write about the election next week, probably Tuesday or Wednesday. The first round is October 4 and the second round is October 25. Lula’s odds of victory are quite high, which might soften the story somewhat as the Lula scenario is already 64% priced in.
2. European energy and political issues in Q1 2027. While the global energy shortage and inventory depletion story has been massively overplayed by most analysts, it’s still real. Remember the winter of 2022, when Europe was running out of fuel and The Economist featured this cover story at the absolute bottom, when EURUSD was trading below 1.00?

I am not an energy expert, but there’s probably a common sense view that Europe might struggle a bit this winter as the heat wave and Hormuz shenanigans leads to the depletion of stockpiles. Here’s a chart from Helima Croft at RBC and a headline from France24 below that. There is not much margin for error going into winter and with gas already expensive and curves backwardated, there is no commercial incentive to fill storage. Starting winter at 70% rather than 80%/90% is fine if winter is normal, but if it’s a mirror of the heat wave insanity, there could be a major issue. This is all just highly speculative food for thought right now.

You have the energy risk, and then you can add some political risk, too. French elections in April 2027 are setting up as a potential risk for France spreads and maybe for the euro. We have seen this movie before, and the market’s good guys always won in the past. Marine Le Pen continues to come back to life over and over, like Mike Myers (the Halloween villain, not the funny Canadian guy).

I often say that themes can only be scary once. For example, the UK fiscal crisis of 2022 is unlikely to replay in the same way because we already know that the BoE and friends will intervene and the thing will end in a flash. Same with tariffs and Iran.

You can only get so scared the third time something happens. Or the fourth. But while this is the fourth time Marine Le Pen has been top of the pops for a French election, the important distinction is that she has never won. So the fear factor should remain alive and well here as we don’t know what a Le Pen-led France would look like.

In 2017 and 2022, Marine Le Pen made the runoff, but there was a powerful front républicain: center-left and center-right voters who disliked one another but still consolidated behind Macron to keep Le Pen (and Rassemblement National (RN)) out. In the 2024 parliamentary election, essentially the same mechanism worked again. More than 200 candidates withdrew tactically to avoid splitting the anti-RN vote, helping prevent an RN majority.

The old story was always that Le Pen can win round one but not round two. Now she might win the presidency. Macron is out so you have a potential Round 2 of Le Pen vs. Mélenchon. That’s not good for markets! The market wants to punish weak fiscal stories (Japan, USA, UK) but has not spent much time hammering France. There are flare-ups here and there but France vs. Germany spreads have not been particularly volatile. Current around 80bps, which is significantly wider than 2013-2020 but nothing super scary. But French debt was 89% of GDP in 2012. 99% in 2017. 114% in 2022. And will be above 120% when the 2027 election happens. It’s ripe for a speculative attack / real money unwind.

To summarize:

The energy story and the France story are both way out on the calendar and fall outside my trading time horizon. That said, if I was running a book that skewed long equities (especially European equities), I would look at the absurdly cheap EUR vol opportunity here. For example:

• 1-year 1.10 EURUSD digital 12% mid
• 1-year 1.05 EURUSD digital 5% mid
• 1-year 19.00 EURMXN digital 14% mid

These trades require some faith and/or imagination. They are worth considering now, when you can put them on with incredible leverage. If and when either of these themes start to pick up steam, vol will be higher and spot will be lower and it will be impossible to put stuff like this on at attractive levels.

I acknowledge this stuff is beyond the time horizon of most readers and involves two squishy catalysts. But it’s an example of “hedge when you can, not when you have to”. Stuff is on sale because everyone is tired and bored.

Final thought

First, you had DeepSeek raising prices. Now you have CRWV saying on their call last night that SKUs from 2020 are still rentable today and renters are signing contracts out to 2029 for those 2020 chips.

CFO Nitin Agrawal said:

“we recently signed an A100 contract that extends into 2029 at an attractive price. As a reminder, this SKU was introduced in 2020…”

A big part of the bear case for CRWV was that GPUs are like smartphones. New generations make old ones obsolete quickly. Therefore, CRWV's six-year depreciation assumption is too optimistic, residual values are overstated, and its true economic depreciation is much higher. In fact, they are now able to rent those 2020 GPUs all the way to 2029 and possibly beyond. Inference does not require Rubin or Blackwell chips and that’s pretty good for CoreWeave’s economics. Just interesting as the AI bear case is getting a tiny bit weaker at a time when sentiment got fairly bearish. Maybe GPUs are more like airplanes, not smartphones?

CRWV bottomed at $60 and it’s trading $107 now, so you are not exactly buying low here, but still. Once a GPU is paid off, the hurdle to make money renting it out is incredibly low. Anyway, the bear case for AI is getting a bit weaker and entry points for many of these stocks have improved dramatically. Long CRWV with a stop below $100 probably works. Target $132 and/or $138. Another suitable expression would be to buy 21AUG $120/$130 call spreads for ~$2.70.

Have a smooth but not criminal day.