• Lykeion
  • Posts
  • NVIDIA's Cracks, Oil's Surge, & Higher Rates for Longer

NVIDIA's Cracks, Oil's Surge, & Higher Rates for Longer

Back to school update on the most important narratives of 2023

First time reading Lykeion? Sign up here.


  • The reality of higher rates for longer is gradually setting in but the market doesn’t seem to care.

  • Oil broke an important technical level and reached the highest price since November 2022.

  • NVIDIA’s 30% consensus revenue growth for the next couple of years has little odds of materializing.

Update on the Key Narratives of 2023

Bad Year for Bonds

So far, it’s been a bad year for bonds, and we might see the third consecutive year of losses for 10-Year Treasury‘s – a historical first.

  • For the first time since Lehman collapsed, 10-Year Treasury Yields have averaged 4% for 50 trading days. 10-Year real rates have also hit the highest level since 2009.

The reality of higher rates for longer is gradually setting in, but the market, for now, seems unfazed by it, with volatility coming down to levels last seen pre-COVID (now might be a good time to re-listen to our chat with Cem Karsan).

After Jackson Hole, the odds of another rate hike (bad for bonds) before November/December increased slightly (still not base case though) thanks to Jerome Powell’s moderately hawkish speech. The overall story of bond yields is nonetheless unchanged – the consensus is that rates need to come down, and they’ll need to come down fast over the next couple of years.

  • For 2024, the market now expects four rate cuts, which, if true, would be a huge positive for bonds.

  • For that to happen though, inflation (what matters, according to the Fed, is PCE Core Services minus Housing) needs to move lower, and with the 4%-5% range being pretty sticky since 2021, there are significant risks that the four cuts next year (previously five) prove to be excessive unless the economy eases significantly.

Strong Year for Equities Despite Macro Outlook

On the US equities side, we’re not too far away from all-time highs, credit due to strong performance from tech and a stronger-than-expected performance from the economy (led by COVID-19 excess liquidity and government-led fiscal spending).

  • A recent series of lunches between Blackstone senior management and high-level investors highlights a bullish outlook: “A majority thought the US economy would have a soft landing. The S&P 500 would hit 5,000 sometime in the next few months. Inflation would move toward 2% but not get there”.

This performance comes against a worsening outlook for real estate, retail sales falling for 9 consecutive months when adjusted for inflation, negative industrial production, the Leading Economic Index declining for 16 consecutive months (the longest since 2007-08), and excess savings forecasted to run out over the coming months.

  • What worries us most is not how inverted the yield curve is, but how it seems to be reverting out of the inversion – something that Roger has flagged many times as the catalyst for a recession to finally materialize.

  • This matters because, as we also flagged many times, the S&P 500 tends to hit a low either during or after a recession takes place, implying that the risk/reward ratio of adding risk at these levels is likely misguided.

We wrote in previous months about how Goldman expected market breadth to widen, but that has yet to take place in a significant manner.

The big air gap in the market to watch for continues to be the performance of tech versus bond yields (usually negatively correlated). No predictions here – the market can stay irrational longer than you or I can stay solvent.

Internationally, the outperformance of non-US markets that took place over the last year or so has reverted in the last couple of months, credit due to a stronger dollar led by the move higher in US rates.

US markets now find themselves at premium valuations versus their 25-year average, whilst the story is slightly different for the rest of the world.

Oil Breaks Its Trading Range

During the summer, we wrote about oil’s underperformance both in our Markets update and especially warned readers in our July Research report of how some strong tailwinds were building for oil in the short term:

  • “Fundamentals are about to turn bullish as stocks start being drawn from June onwards as demand outpaces supply by 2.4 million barrels per day on average in 2H, driven by OPEC+ cuts, global aviation recovery, and China’s economic recovery.

  • The last two times the market was this undersupplied, oil moved from $51 to $75/barrel (up 47%) between 2017 and 2018, and from $44 to $69/barrel (up 57%) between 2H20 and 1H21.”

Since then, oil has decisively moved higher but, most importantly, it has broken an important technical level: the 50DMA crossed the 200DMA – what’s known as the golden cross. The last time that happened, oil entered a bull market that lasted almost 2 years and got everyone worried about energy scarcity.

We’ve also had, for the first time since we started draining the SPR, four consecutive weeks of reserve increases.

Additionally, Saudi Arabia and Russia extended their voluntary cuts through year-end, leading oil to its highest value since November last year.

  • Since mid-2022, the energy sector has decoupled from the volatility of oil prices and is now about to break all-time highs.

Despite that, it’s still the sector in the US market that trades at the largest discount to its 20-Year PE average – by a wide margin. We’ve said it many times, but the oil supercycle is here to stay and there’s still plenty of upside left.

AI Boom Show’s No Signs of Bust

From The Information: "OpenAI is currently on pace to generate more than $1 billion in revenue over the next 12 months from the sale of artificial intelligence software and the computing capacity that powers it. That’s far ahead of revenue projections the company previously shared with its shareholders, according to a person with direct knowledge of the situation". OpenAI made $28 million in revenue last year.

These kinds of headlines are continuing to add momentum to the AI bubble that started earlier this year.

NVIDIA is close to all-time highs again (despite yesterday's move lower), now trading at 17x next year’s sales, with revenue expected to grow annually by 30% for many years to come.

  • Bringing back Scott McNealy’s quote, the billionaire businessman who co-founded Sun Microsystems, a darling of the late 1990s that crashed during the tech bubble implosion, seems timely here:

  • “We were selling at 10x revenues. At 10x revenues, to give you a 10x payback, I have to pay you 100 percent of revenues for ten straight years in dividends. That assumes I have zero cost-of-good-sold, which is very hard for a computer company. That assumes zero expenses, which is really hard with 39,000 employees. That assumes I pay no taxes, which is very hard. And that assumes with zero R&D over the next ten years, I can maintain the current revenue run rate. Do you realize how ridiculous those basic assumptions are? You don’t need any transparency. You don’t need any footnotes. What were you thinking?”

To be clear: living up to 17x sales is not impossible, but it is a significant hurdle to overcome. How high of a hurdle you might ask?

  • Over the last 30 years, only 7% of FTSE World Index (3,400 companies) have sustained revenue growth of more than 30% per year over a five-year period (leading to a quadrupling of sales).

  • For larger companies (Top 200 of the index) the feat is even rarer: only 45 businesses ever achieved that growth over the last 30 years.

  • The picture looks slightly better when filtering for valuation multiples: 23% of large businesses trading at least at 10x sales ended up delivering 30% per year revenue growth over five years over the same time frame.

  • But this still means that only one of four companies that trade above 10x sales end up meeting growth expectations.

  • To be clear, to make money out of NVDA, you can’t just meet expectations (that’s already in the price) – you need the business to surprise to the upside. And for a stock that has fallen by more than 50% more than 10 times and by 80% three times in its history, it would be foolish to expect smooth sailing from here onwards.

And to close this one out, there’s a growing narrative out there in social media (still to be properly picked up by financial media) that highlights how NVDA might have been manufacturing GPU demand… if this is true, gravity should come in and hit hard not only NVDA but the market overall. Recommended read.

If you’re not already signed up for our Research tier, here’s a nudge to give it a shot. This is where we break down the frameworks that help us inform about current opportunities and risks within macro and geopolitics. Last month we focused on:

Thanks for reading through! Obviously, none of this is investment advice.

As always, we'll see you out there...