Take a Hike

Edward Lim, _CFA

September 2026

At the start of the US-Soviet Cold War, President Dwight Eisenhower delivered a speech at the National Defense Executive Reserve Conference in Washington. He was addressing largely civilian leaders, urging them to prepare their companies and their personal faculties in the event of a prolonged Cold War. The timing of the speech was particularly tense because, just weeks earlier, the USSR had launched Sputnik 4, intensifying fears in America that the Soviet Union was overtaking it in space and posing a significant military threat. In that speech, he said, “Plans are worthless, but planning is everything,” and “if you haven’t been planning, you can’t start to work, intelligently at least.” Managing investment portfolios is all about planning. Sometimes the plan turns out to be wrong, but in planning for a particular course of action, one has inadvertently also considered alternatives.

In our last Navigator, World in Motion, we stated that even as the global economy slowed due to rising oil prices, it would remain resilient, supported by the trifecta of an AI spending boom, the lift it provides to many parts of the economy, and very healthy corporate profits. Inflation would remain troublesome but, on balance, a weakening labour market would prompt the Fed to cut as its next move, opting to support “maximum employment” over “price stability”. The first two propositions have held true, but the third, on Fed to ease, has not. On 16 September, the Fed raised its target range by 25bps to 3.75–4.00% after holding rates for nine months. And as President Eisenhower said in the same speech, when your planning is wrong, “the first thing you do is to take all the plans off the top shelf and throw them out the window and start once more.”

The wrestle here is that while the economy has strengthened, so has inflation, and it is now posing a bigger concern than the languishing labour market. At the micro level, earnings estimates are rising and equity valuations have become more reasonable. Yet the prospect of further monetary tightening threatens this four-year-long bull market. Three decades in this industry have humbled me enough to respect the maxim, “Don’t fight the Fed”. We retain a neutral allocation to equities and an underweight in bonds, look to rebuild our gold holdings, and retain meaningful exposure to hedge funds.

Resilient growth keeps the bears at bay

Our three familiar lenses, nowcasting, purchasing managers’ surveys and consensus forecasts, not only continue to point away from recession, but also towards a global economy that appears to be re-accelerating. The near-term recovery has been stronger than the longer-term forecast revisions might suggest. Nevertheless, the downgrades in longer-term forecasts since the war broke out have also stabilised for most regions.

The Nowcaster put annualised global growth at 3.3%, an improvement of 0.3 percentage points from our previous quarter’s update and well above long-term potential. Developed markets have strengthened to 2.4% and emerging markets to 4.6%. The US is running hot at 3.3%, while China has stabilised at 5.1%. India and Brazil are exceptions to the improvement in momentum, although slower growth should not be confused with contraction.

Nowcast growth momentum has improved

Well above long-term potential as well

August’s global PMI rose for a fifth consecutive month to 53.5, consistent with approximately 3.1% annualised growth. Manufacturing remains supported by AI investment, while services have recovered from their earlier softness. Expansion is broad, with 74% of tracked manufacturing PMIs above 50 and 71% having improved from a year earlier, even as month-on-month momentum is choppy.

Services have recovered, manufacturing remains expansionary

Breadth of expansion remains healthy

Looking further out, consensus forecasts put global growth at 3.0% in 2026 and 3.1% in 2027, with the US at 2.1% in both years. Emerging-market forecasts have suffered larger downgrades than those for developed economies, stemming from weakness in the GCC and in countries sensitive to imported oil, such as India. Nonetheless, the aggregate picture remains one of expansion.

Inflation has become less obliging

Unfortunately, the improvement in growth comes against the backdrop of another surge in crude oil and downstream products, from diesel to fertilisers, while renewed supply-chain pressure has all but reversed the disinflation trend seen in the first half of the year. Take the US as an example. Monthly readings show renewed pressure across both headline and core inflation, including goods and services. A central banker can look through a temporary increase in petrol prices. It becomes harder when the pressure spreads to the rest of the shopping basket.

Near-term inflation problem extends beyond energy

We still think there is a credible case for inflation to ease towards the Fed’s target in 2027 as the effects of tariffs wane, wage growth slows, and the rent components of CPI continue to ease, consensus forecasts put US CPI at 3.3% this year, falling to 2.4% next year. There are also one-off adjustments in 1H26 CPI numbers such as higher financial-services and software pricing that will not be present in future computations. We will keep marking the four main variables driving inflation: oil, goods prices, wages and rents. The first two have become less comfortable, but the latter two still offer some relief.

Wages and shelter provide some restraint on higher inflation impulses

Oil has used up some of its buffers

In our last publication, World in Motion, we explained why the initial oil shortage had proved less destructive than feared. Consumers adapted, supply outside the Gulf increased, throughput through the Straits was greater than feared, while inventories absorbed the shortfall. The oil-balance model now estimates that the global deficit narrowed from about 7 million barrels a day in March to around 1 million in the third quarter. That is a substantial improvement. But the stockpile that made this adjustment possible is depleting. Visible global inventories have fallen by more than 500 million barrels since March, towards the lower end of the available historical range. Measured against demand, cover has retreated towards 74 days, leaving the supply chain less able to absorb another interruption.

While the deficit was less than feared, the inventory cushion has shrunk

The composition of oil inventory matters too. The significant drawdown in strategic reserves and oil on water helped bridge the disruption, but both are now close to safety-level thresholds. The remaining source of inventory is OECD commercial inventory. That has remained broadly unchanged and close to its recent average, but these are barrels available at a market price. Their owners have little reason to provide the same insurance on the same terms as an emergency reserve release. The combination of uncertain flows and thinner inventories raises the risk of another price shock. The economy survived the first disruption better than feared. Repeating the exercise with less buffer would be harder.

Gulf exports remain vulnerable

Big draws on SPR and oil-on-water inventory

Warsh says take a hike (or two), and so did many others.

Our earlier expectation of a cut placed considerable weight on labour-market softness and the Fed’s willingness to look through a supply shock. That balance has shifted. Stronger activity gives the Fed more room to confront inflation, while the persistence of price pressure makes waiting untenable. The latest dot plot suggests another hike by the end of the year, which we think will likely come in December to avoid accusations of politicization. It has also removed the previous projection of a rate cut in 2027. We lean towards a shallow hiking cycle, with subsequent decisions depending on the inflation mix and developments in energy markets. Elsewhere, the share of central banks whose last move was a hike has risen to 41.2%. That is a meaningful turn towards tightening, but still far from the last tightening cycle in 2022–2023.

One more hike in 2026 and no cut in 2027

41% of global central banks hiked in the last 2 months

Asset Allocation Strategy

Equities: Still neutral in the short term, but interesting long-term entry points are emerging.

With the Fed moving again, it is tempting to reactivate muscle memory of what happened in the last tightening cycle of 2022–2023 when holding neither equities, bonds nor even gold made you money, let alone provided diversification from one another (MSCI World Equities fell 20%, Global Bond 17% and gold 1%). History offers a more discriminating answer. We examined twelve first-hike episodes from 1971 to 2022. In the short term, in ten of the twelve episodes, the S&P 500 was lower three months later. The average decline was 4.1% and the median 2.1%. It does not help that we are also running into midterm elections, which have tended to be negative for equities as well. Each on its own is sufficient reason to respect the near-term risk.

Equities: the first three months were difficult. Beyond that, the probability of positive returns rises

However, the longer horizons are more optimistic. After six months, outcomes were evenly divided and the average return was approximately flat. Nine months and a year out, the probability of positive returns improved to 67%, with mean returns of 3.0% and 5.6%, respectively. When we distinguish returns in a fast-hiking cycle from those in a slow/non-cycle (fast cycle is defined as the Fed hiking at each meeting versus slow cycle across a longer schedule of meetings and non-cycle are often), outcomes differ materially. As we are in the camp that this is likely a mid-cycle hiking adjustment, with subsequent decisions characterised by data dependence, history points to positive returns after a first hike, with the return profile improving as the months pass (first month: +2.2%, sixth month: +3.8%, ninth month: +7.5%, twelfth month: +17.6%).

In the current environment, we think what matters more for equities is less about the Fed’s pace of tightening and more valuations, together with the risk of overestimating earnings. As shared by my colleague in Equity Views 2Q26: The AI Debate: Separating Fundamental Momentum versus Price Momentum, he laid out the case for strong AI adoption and the phenomenal economics of AI compute, which have led to a continuous increase in hyperscalers’ capex. As earnings uplift has outpaced price increases, the valuations paid to own these stocks have actually fallen. Take NASDAQ as an example. It has risen 14% to date, yet its 2026 valuation multiple has fallen from 28.4x P/E at the start of the year to 19.2x currently because 2026 earnings growth has been revised from 21% at the start of the year to 36%. This is not confined to US markets. On a rolling 12-month forward P/E, many markets are now trading around their historical averages in contrast to the past four years when most were trading at or one standard deviation above their historical averages because EPS estimates have been revised higher.

Valuation multiples have contracted as EPS growth and revisions have outpaced price increases

2026/27 have seen extraordinary +ve EPS revisions

Quarterly momentum remains strong till 1Q27

While we may not have a valuation bubble, there is always a risk of an earnings bubble if AI-related revenue estimates prove too optimistic because use cases and ROI arrive more slowly than expected. This AI supercycle is already delivering EPS growth faster than its historical long-term trend. In 2026, hyperscaler capex (4ppt) and AI infrastructure spending (15ppt) are the main drivers of 28% growth, implying the rest of the S&P 500 grows by only 10%. The same dynamic applies in 2027: of 20% forecast growth, 11ppt comes from AI semis and infrastructure-related companies, and 3ppt from the main hyperscalers, together accounting for 70% of next year’s growth. This is where equity vulnerability lies.

Over-earning in this AI cycle?

AI earnings drive much of S&P growth

We commented in our start-of-year Navigator, Bubblicious, that we were at the cusp of enterprise adoption in 2026. When Anthropic launched Codex for Claude in March, nobody would have guessed that in less than nine months, Anthropic would achieve over $70bn of annualised revenue run rate by simply producing better software code. In comparison, Microsoft, the largest software company in the world, generated total software revenues of $268bn last year. A company less than three years old is now generating revenue equivalent to one quarter of that of a mega-cap tech firm with 51 years of heritage. As it stands, hyperscalers’ backlog now stands over $1.6trn, a marked jump from the start of the year of $600bn. Our team has just returned from Korea, and the trip confirmed our belief that physical AI has arrived on the manufacturing floor. The recent breakthrough by Moderna in developing a personalised mRNA cancer therapy with Merck is another validation. The CEO of Moderna said, “If we had to do it the old biopharmaceutical ways, we might need a hundred thousand people today. We really believe we can maximize our impact on patients with a few thousand people, using technology and AI to scale the company.”

At our dinner event held in July 2025, I mentioned the concept of sovereign AI, arguing that it would become a national imperative for countries to have native LLMs and data centres located within their own borders. Sovereign AI is central to a nation’s competitiveness, strategic resilience and institutional trust within society. It must be built and controlled by national bodies, not corporate America alone. The TAM of $200bn by 2030 estimated by Goldman Sachs back then has since been surpassed by the latest report from McKinsey, which expects it to become a $500–600bn market over the same period and account for 30–40% of total global AI spending. This year alone, Korea and Japan have announced multi-year plans of over $600bn each, India $200bn, while Singapore has a national AI R&D plan of $1bn. We continue to think the market will be surprised by the trajectory of AI capex, and that the next group driving this capex will not just be the hyperscalers, but also the neo-cloud providers; not just closed source but also open-weight models.

Fixed income remains underweight: Treasury yields almost always move ahead of the first hike. Prospect of lengthening duration soon.

Rising inflation remains a headwind for bonds, but interestingly, US Treasury yields often rise ahead of the actual hike as markets build expectations of an impending Fed move. On average, US 10-year yields rise 45–58bps in the months before the first hike. In the post-first-hike period, the odds are even across different time horizons, while a slow-hiking cycle usually sees Treasury yields decline instead.

Bond markets often move well before the first policy hike. Post-hike reaction is more subdued

Another feature of the post-first-hike period is that credit spreads tend to narrow, reflecting strong economic activity, while the yield curve generally flattens. On balance, our long-standing underweight in bonds becomes less compelling the further the cycle tightens while the economy remains robust.

Credit spreads tighten most of the time post-first-hike

Alternatives: Hedge funds are a better diversifier for equities than bonds.

We retain meaningful exposure to hedge funds because the next policy surprise should not require every part of the portfolio to behave in the same fashion. The experience of 2022 remains an expensive reminder of what happens when inflation makes both bonds and equities vulnerable to the same repricing risk. Our fund of hedge funds, GARP, is intended to reduce that dependence through managers operating across different markets and strategies and generating returns that are uncorrelated to the broader equities and movement in yields.

Commodities: Gold deserves another look, Bitcoin halving overhang nearly over.

In recent quarters, we have been reducing gold and bitcoin as rising yields and a stronger dollar made the balance of risks less attractive, given their competition with interest-bearing assets. Our intention now is to rebuild towards a 5% allocation following the September policy decision. Inflation uncertainty, depleted energy buffers and the risk of further geopolitical disruption strengthen the case for holding this insurance, particularly gold. In the same historical study, both the average and median returns were positive at the three-month and one-year horizons for gold. Moreover, we have seen a significant reduction in speculative positions in both assets over the last nine months, and for gold, the recent renewal of central-bank buying reinforces its investment case in a multi-asset portfolio. Spurious as it may be, given bitcoin’s limited trading history, the data suggest we are near the end of the decline in the second year after each halving. If history is a guide, bitcoin winter could end by 4Q26. Furthermore, we have also noticed that with each halving cycle, the duration and extent of the drawdown in this asset class have diminished, perhaps reflecting greater institutional participation.

Bitcoin halving near the end?

Cash holding is larger than usual: 5–15% for flexibility

Short-term Treasuries remain a useful holding while we assess the next move in inflation and rates. They provide income and flexibility to add risk when prices become more attractive.

The portfolio will change again if the evidence warrants it. Sustained moderation in goods inflation, more secure Gulf flows and continued earnings upgrades would improve the case for adding to equities and extending bond duration. A renewed energy shock, accompanied by weaker margins and deteriorating revisions, would lead us to reduce risk.

Featured Picture/Quote:
“SACRIFICE__YES_if_you_accept_permadeath,”
OpenAI agents going kamikaze

Edward Lim, _CFAChief Investment Officer
edwardlim@covenant-capital.com

Risk Disclosure

Investors should consider this report as only a single factor in making their investment decision. Covenant Capital (“CC”) may not have taken any steps to ensure that the securities or financial instruments referred to in this report are suitable for any particular investor. CC will not treat recipients as its customers by their receiving the report. The investments or services contained or referred to in this report may not be suitable for you and it is recommended that you consult an independent investment advisor if you are in doubt about such investments or investment services. Nothing in this report constitutes investment, legal, accounting, or tax advice or a representation that any investment or strategy is suitable or appropriate to your circumstances or otherwise constitutes a personal recommendation to you. The price, value of, and income from any of the securities or financial instruments mentioned in this report can fall as well as rise. The value of securities and financial instruments is affected by changes in a spot or forward interest and exchange rates, economic indicators, the financial standing of any issuer or reference issuer, etc., that may have a positive or adverse effect on the income from or the price of such securities or financial instruments. By purchasing securities or financial instruments, you may incur above the principal as a result of fluctuations in market prices or other financial indices, etc. Investors in securities such as ADRs, the values of which are influenced by currency volatility, effectively assume this risk.

Back to Perspectives

Take a Hike

September 2026

Disclaimer

Accredited & Institutional Investors

Personal Data Protection Act

Terms of Use

Privacy Statement