Macro Pulse for October 19: Observations from the Road

• Concerns around private credit quality are mounting as cases of fraud pop up globally. Meanwhile, U.S. banks have tapped the Fed’s emergency short-term lending facility for $15B over two days, suggesting liquidity pressures in the repo market. We do not believe either development represents the “tip of the iceberg” of systemic issues, particularly as the Fed eases, but see credit discipline increasingly rewarded as we enter later a later stage in the cycle.
• The Fed is overwhelmingly likely to cut in its October 29^th meeting, supported by risks of a credit or liquidity crunch, the known halt in U.S. hiring, and the marginal negative impact to the economy from the U.S. government shutdown. As the shutdown drags on, we look to the sole exception of the pause in economic data releases: the September CPI report, originally due out October 15^th but postponed to October 24^th. We expect anything other than an extreme upside surprise in inflation to cement a further Fed cut.  

Is AI in a bubble?

Most investors are convinced that AI – and the buildout of its infrastructure and application layers – are here to stay. AI has become a central economic and market driver, and as Sarah and I discussed in yesterday’s Macro Pulse webinar (replay here), the U.S. and global economy have AI to thank for how stable aggregate investment figures have been, given uncertainty in other areas.

But high valuations have led to a considerably more skeptical shift in tone.  Has enthusiasm for AI gone too far? In Europe, our conversations centered around the following potential risks. GMS will share our own vetting of these risks in a note in the coming weeks: 

• High valuations increase the risk that an accident in a relatively small set of AI “winners” disrupts a top-heavy market for everyone (see: Cisco in 2000). This is important not only for the market but for the real economy. The top 10% of income earners are now contributing close to 50% of the consumer spending in the U.S. A deterioration to the wealth effect supporting that spending would impact everyone.
• AI’s electricity demand has already increased power costs globally. Investors are questioning whether power generation improvements can keep pace with AI, or whether we’ll see a near-term disruption, which could stem from a lack of supply or from regulation intended to protect households from rising electricity costs (see: the NJ governor race).
• The circularity of money spent on AI is increasing. A recent example of this is the deal between OpenAI and Nvidia: Nvidia agrees to invest $100 billion in OpenAI, which then buys a lot of Nvidia chips with cash it gets by selling Nvidia stock. Circular flows of money may represent supply chain integration typical in other industries, or could be a warning sign that capacity is expanding too rapidly.
• Financing of AI infrastructure is shifting. For the first two years of this trend, companies financed their AI capex buildout mostly from cash on their balance sheets. That’s now transitioning to debt financing, creating a bridge between a cash-driven cycle and a credit-driven cycle. 

Why all the attention on gold?

Demand for gold is increasing because the global economy is in the middle of a critical geopolitical transition.

Central banks began exchanging dollars (and, at the margin, euros) for gold more than a decade ago, but that accelerated in 2022 when the U.S. froze Russia out of SWIFT. This structural force is unlikely to reverse.

Central bank backing of gold’s runup is getting more attention now – and therefore prices are moving higher, for a few key reasons:

• Political risk: relative instability of U.S. policy has intensified the transition towards currency hedges, including with gold.
• Supply: new mines are hard to come by, so supply has been limited.
• Inflation: inflation risk lingers due to AI capex, tariffs, and other risks. Gold provides a 50-year proven track record as a hedge.
• Momentum: as prices go up, investors want in; gold is rapidly becoming a “FOMO” trade.

We’ve been writing for a year that gold should have a greater place in portfolios. Bridgewater founder Ray Dalio recently said “investors should be allocating as much as 15% of their portfolios to gold.” While we may not agree with a single 15% position in gold, we think investors should hold up to 10% in real assets, including gold and other commodities.

U.S. assets are back in favor, but diversification is “real” now.

Diversification has long been a central tenet of portfolio construction. But if we’re honest, the last 15 years’ ideal asset allocation didn’t need it. Throw all your money in U.S. large cap growth equity and call it a day.

We like U.S. large cap growth equity, but there’s no doubt that the critical risks to portfolio construction have risen along with valuations.

As a result, we’ve seen global investors take diversification more seriously:

• Bringing underweights to ex-U.S. equity toward neutral
• Taking small cap exposure in the private markets
• Adding serious satellites to gold, precious metals, industrial metals, and other real assets
• Neutralizing duration while maintaining risky bond exposure

… and focusing more on security selection, especially in credit. This point sounds obvious but I want to underscore it: on a cross-asset basis, it hasn’t necessarily paid to be disciplined as the market rallies. We are still fully allocated to credit but believe selection and conservatism become more important as the cycle matures.

And on that note – should be we be worried about private credit?

Developments in Tricolor, First Brands, Zions Bancorp, and Western Alliance have accelerated questions about the credit markets in general, and private markets in particular.

Each of these credit events involve the discovery of fraud, and fraud is not typically systemic. However, the market has been pricing not only the fraud events –dragging broader financials with it – but also (in our view) a concern that these events represent the tip of the iceberg: the dollar has weakened, the 2-year Treasury has moved lower.

The market may also be sniffing out that the Fed will have to be more supportive of the economy given these developments. A bank (or banks) used the Fed’s standing repo facility for two days straight last week, a relatively rare occurrence suggesting illiquidity in the money markets. We would argue, thanks to calm SOFR spreads, that Fed facilities have done exactly what they’re designed to do. But we’ve already been expecting the end of the Fed’s quantitative tightening program at some point in 2025, and may very well see that announcement in its October meeting. A second additional rate cut this year may also become more likely in light of a real – or even perceived – liquidity crunch.

We do not believe a financial “blowup” is imminent, but we do believe we are moving later in the credit cycle. Credit costs have moved meaningfully higher in recent years, bringing in more capital to the space chasing too few deals. It’s not unusual for that kind of deal pressure to result in declining credit discipline, allowing for select cases of fraud to go unnoticed. 

As I mentioned above, credit discipline hasn’t consistently paid since the GFC. We’d now argue it will. A Fed cutting cycle into a still-resilient economy should pull money market cash off the sidelines and into the credit markets, as long as confidence in overall credit quality holds up. We believe that staying invested in credit but with a heavy focus on manager and security selection is prudent.