Fading the Utes
In Section 2 of this week’s Monday notes, we revisited the last of Grandpa’s Defensives from our April note. The utilities. The AI party has arrived at the door of the market’s most boring sector. While I completely understand that power demand for data centers running the Large Language Models of generative AI, market extrapolation of the narrative appears to have gone wild.
On April 22nd we wrote:
“You already know that the 🐿️ is no bond bull. ‘Bond proxy’ utility equities offer the upside of regulated returns and come with a heap of baggage. I am beginning to hear whispers of a bullish thesis for electricity producers around AI data center related power demand. Would love to be contradicted, but I am not sure that I get how (mainstream) utilities can effectively capitalize on this trend.
Sure, our West Coast based tech overlords have finally woken up to the fact that their new AI toys are going to require more electrical power than the grid has to offer them. This is only going to place additional strain on an industry that is already facing numerous challenges:
- Demand - potentially tripling by 2050. Electric vehicle penetration forecasts may now be looking less like hockey sticks, but AI data center power demand is a whole new ball game!
- Ageing power production assets and a grid in desperate need of overhaul and hardening - just as the cost of debt and equity is rising;
- Utilities are consumer-facing businesses. They too have ‘political football’ risk in a populist world (remember that they came after the ‘price gouging’ petrol/gas station owners who make most of their money from selling you coffee and candy!);
- Political / ESG decarbonization targets that have diverted investment away from core infrastructure towards less reliable wind and solar renewable capacity;
- Increased heat and drought conditions. Climate is already creating litigation liabilities for utilities (wildfires in California and Hawaii), now power companies (particularly in the Western US) must deal with disruption from water stress and summer cooling (AC) demand.
There may be individual examples of independent players that can manage the regulatory red tape and capitalize on the opportunity to meet the demand from AI’s hungry power users, but at the end of the day don’t the tech overlords just end up building these capabilities themselves (probably with SMR nuclear technology)?
XLU 0.00%↑ the SPDR Utilities ETF is up a further 12% since that note. The 🐿️ can ignore no longer!
I am happy to concede that certain players, such as the IPPs, may selectively benefit from increased baseload power demand. I further concede that this is good news for ‘Utes’ with a significant fleet of nuclear assets. However, the majority of the players in the US utility market are governed by regulated returns on their regulated asset bases.
Jacob Shapiro and his partner Rob chatted about this on this weekend’s edition of the Cognitive Dissidents podcast. Please do listen to the whole show, but I found myself nodding vigorously (from minute 30) when Rob was discussing this same point on the ‘Utes’. He also gives an elegant explainer on how to think about regulated asset returns.
Repeat after me. Utilities are not AI stonks! Yet the sector as a whole is running like it stole something! So much for bond proxies, the sector’s relationship with bond yields (traditionally reliable) has completely broken down since early this year. The 🐿️ needs to fade this.
In the past 3 months, the XLU 0.00%↑ utility ETF, has returned almost 20%. Half of the contribution has come from the IPPs and the ‘nuke-heavy’ constituents of the ETF, but the whole sector definitely feels somewhat over-cooked at this stage.
Arjun Murti’s excellent piece this week also touched on this as well as the merchant power opportunities among the US natural gas producers. I completely agree that there are interesting opportunities developing here. However, some of the Natural Gas producers have been doing some AI performance art of their own recently!
How to play it
Regular readers will know that we have been staying away from the ‘dark side’ (shorts and puts) in equities for quite some time. To be clear, I am not looking for a massive correction in the sector. However, a gentle mean reversion / gradual letting of air out of the tires does seem appropriate.
We are going to express this view with a modified (broken wing) butterfly strategy, with the body of the structure struck in the middle of that long term trading range ($63 per share). The structure has a highly convex payoff in the event that a precise retracement to that level occurs.
The structure can be put on for a net credit of $1.62 per option combination (link to OptionStrat worksheet):
Long 1 x January 17th, 2025, $55 Call (Premium per option: $18.03, Delta 100%)
Short 2 x January 17th, 2025, $63 Call (Premium per option: $10.75, Delta -200%)
Long 1 x January 17th, 2025, $77 Call (Premium per option: $1.85, Delta 40%)
It can also be managed in a way that ensures that any losses on the structure are minimal. We plan to close or restrike the position in the event that the ETF appreciates against us. The hand drawn pink line on the image below illustrates the breakeven point of the structure over time.
The structure is not especially sensitive to changes in implied volatility but will benefit marginally from a pickup in volatility from these (very suppressed) levels.
We will enter the position today and track it as an Acorn. Will post a revised worksheet with final execution levels in The Drey.
If you act on anything provided in this newsletter, you agree to the terms in this disclaimer. Everything in this newsletter is for educational and entertainment purposes only and NOT investment advice. Nothing in this newsletter is an offer to sell or to buy any security. The author is not responsible for any financial loss you may incur by acting on any information provided in this newsletter. Before making any investment decisions, talk to a financial advisor.