One week / one topic: Vintage tomorrow

What happened?

It is no secret that government bonds have been in a funk for a while, especially when compared to seemingly invulnerable stock returns.

­Adding insult to injury, government bonds have been experiencing volatility comparable to stocks… but without anything even remotely similar in terms of returns.

Put differently, government bonds have been behaving like a risk asset… have they conclusively moved from ‘risk-free asset’ to ‘return-free risk’?

And yet, the current debate between bulls and bears reveals that some important developments are probably being overlooked.

As of today, I would summarize the embedded disagreement between the two camps as follows:

• Bull case: We are in a structurally supportive nominal growth regime, and markets are correctly looking through short-term noise.

• Bear case: Markets are pricing a benign equilibrium that is internally inconsistent — and the weak link is rates.

While it’s hard to conclusively argue in favor of either point of view (welcome to markets!), I would still tease out two relevant observations.

First – in the face of a noticeable rise in long-term bond yields – the point of contention is mostly centered around this question: are higher yields a sign of strength… or the tightening mechanism that breaks the system?

Bulls would argue that higher yields are driven by growth and investment demand, and not just inflation, and that real yields reflect productive capital demand for AI and infrastructure.

Bears would counter that term premia are rising structurally because of fiscal supply, runaway AI capex financing and permanently higher inflation volatility.

Secondly – while using the word ‘bubble’ can be damaging, as you witness prices running away for longer than you think possible – there are clear signs of excess.

My favorite is probably the AUM raised by a single-stock, leveraged ETF tracking the performance of SK Hynix, a Korean semiconductor manufacturer.

Granted, it’s hard to sit on your hands while 78,000 Samsung employees are getting a bonus of $340k each. But… from zero to $10bn AUM in seven months? Really?

In summary, visible extremes might argue in favour of fading runaway stock rallies and leaning into unloved govies.

But what’s really at play here?

Our observations

• Fundamentals: Focusing on the Nasdaq 100 as representative of the current ‘AI or bust’ fixation, we can observe it has been trading at significantly higher valuation multiples since the arrive of chatGPT.

  Maybe this is excessive (or maybe not enough?), but at what point does overextended positioning take over and completely trump fundamentals? Does it even have to?

• Price action: As previously noted, markets have remained selective by clearly differentiating between (perceived) winners and losers.

  Some of it feels too obvious, however. Yes, we’ll need more power, memory and semiconductors to burn at the altar of the new digital gods. But human ingenuity has historically displayed a recurring tendency to solve scarcity issues…

• Investor beliefs: The Iran war has notably moved to the back burner, based on assumptions that a deal will eventually be reached and we’ll return to something close enough to business as usual.

  AI enthusiasm on the other hand shows no signs of abating, with new stories and sensationalistic claims every week. Can something as nerdy as ‘sticky inflation concerns’ spoil this party?

So what?

Going back to the core observation that higher yields remain central to the current debate across markets, we should then focus on this question.

For simplicity, one tends to own government bonds for largely two reasons: generating returns, and acquiring portfolio insurance vs periodic equity drawdowns.

As for the latter point, we should then focus on verifying the following hypothesis: do government bonds still provide portfolio insurance? Do yields still go down when growth is softening?

Since (thankfully) we don’t get recessions every year, we can’t easily reach a watertight conclusion.

By lowering the bar, though, we can observe whether yields fall in the presence of ‘textbook’ negative news such as weakening PMIs (Purchasing Managers’ Indices), employment or growth data.

Longstanding frameworks about diversification assume that bonds should zig when equities zag, but – in reality – there have been decades when this was not the case.

It would be ironic if – in the middle of the AI revolution – we returned to an equity-bond correlation regime last seen in the 1990s… or even earlier.

Mood music: Daft Punk – Touch

By popular demand, here is the One week / One topic playlist

The information provided should not be considered a recommendation to purchase or sell any particular security.