Why Rebound Investing Works

There’s a popular saying in finance that goes:

While the Iran crisis is far from over, if you were “monitoring the situation”, it felt like the end of the world. Not to name names, but even investors that we looked up to were panic-selling and acting like we are the beginning of World War 3.

Yet just a month later, the S&P 500 is at an ATH, and the quality names investors sold out during the crisis have rebounded strongly.

This is not just a one-time thing. It’s a repeating pattern.

2020 — Covid caused Rolls-Royce stock (ADR) to drop <$1

2022 — Crypto crash caused a 66% drawdown for Nvidia

2025 — ASML dropped 35% due to China restrictions by the U.S.

Fast forward to today, Rolls-Royce is now worth $150B, Nvidia is 16x from its 2022 lows, and ASML has doubled in the last year.

This is the entire thesis behind Rebound Capital. Buying great businesses when the market is irrationally (or rationally!) afraid. Every once in a while, the market will give a window of opportunity to load up on quality companies. We try to take advantage of the same (we are agnostic as to what presents us with the opportunity - whether it is macro or idiosyncratic events).

And it’s not like we just talk about these companies - we put money where our mouth is and doubled down on quality names during the latest crisis.

The result: while the S&P 500 recovered approximately 10.8% from its March 31st, our portfolio recovered 13.5% over the same window.

That 350 basis points of outperformance doesn’t matter (as it's short-term). But what matters is why it happened, and the names in the broader market that exemplify the rebound pattern.

We present a snapshot of ~50% of our portfolio below, with the % rebound for the stocks (vs. ~10% for the S&P 500). These are point-in-time figures, but we estimate the following stocks can keep compounding from here. The idea behind showing the sharp rebounds is to present a live case study of rebound investing. There are other names like Nvidia, which we took a position during the war, and which rebounded ~20% from its March lows.

Amazon has rebounded 25% from its March lows. Even before the war started, Amazon's shares were weak. The stock had fallen 10% after Q4 2025 (5th February) earnings on fears that the $200 billion capex commitment would fail to generate adequate returns. We strongly disagreed with that thesis then, and all signs point towards the fact that we were right. Read our original deep dive here.

AI use cases are proliferating faster than even optimistic forecasters expected. Anthropic has reached $30 billion in annualized revenue as of April 2026, up from $9 billion at the end of 2025 - 3x growth in 3 months, with no precedent in B2B software history.

Every dollar Anthropic earns runs largely (by RC estimates) on AWS infrastructure and Trainium silicon. Meanwhile, compute scarcity is real: Trainium 2 is nearly exhausted, Trainium 3 is almost fully booked, and a significant portion of Trainium 4 is already reserved. A substantial portion of that capex has committed customer backing — including OpenAI’s $100 billion+ commitment to AWS. AWS AI services carry over $15 billion in annualized revenue; its custom chip business exceeds $20 billion, growing at triple-digit rates (2025 Amazon Shareholder Letter).

Crucially, AWS is a neutral cloud provider that doesn’t compete with the frontier labs it serves - a structural advantage the market isn’t fully pricing in. Both Google and Microsoft will need to compete directly with the AI labs (Anthropic and OpenAI) to protect their core business.

AMD dropped 17% in February 2026 after Q4 earnings, where guidance overshadowed a strong beat: the company reported $10.27 billion against consensus of $9.67 billion, guided Q1 above analyst expectations, and still got punished. A classic overreaction. The market just got impatient, and the sentiment was weak.

Read our analysis on AMD here.

We bought AMD because after the selloff, it was trading at roughly 10x its medium-term earnings power. At AMD’s November 2025 Analyst Day, Lisa Su, who has a well-established track record of under-promising and over-delivering, guided to 35% annual revenue growth over three to five years, with the AI data center growing at 80% per year.

Since then, OpenAI committed to deploying up to 6 gigawatts of AMD Instinct GPUs (this deal was public before the Q4 earnings overreaction), Meta followed with its own 6-gigawatt agreement, and reports of a potential Anthropic collaboration would further cement AMD’s position as the dominant second-source supplier in the AI compute arms race. We estimate that AMD’s revenue by 2030 can be 2x -3x larger than today.

CSU fell by more than 50% from its highs amid fears that AI would disrupt its software businesses, fears that reveal a fundamental misunderstanding of CSU’s moat. Constellation acquires and operates vertical market software for niche industries: funeral homes, transit authorities, and agricultural co-ops. The moat was never code generation (which AI solves for); it’s decades of workflow integration, institutional trust, and regulatory entrenchment. These aren’t categories that attract AI startups because the TAMs are small and sales cycles are long.

On valuation, CSU’s FCF yield has reached approximately 5% - rarely seen for a business compounding at double-digit rates. Headline earnings are depressed by acquisition-related amortization. CSU has virtually zero SBC (stock-based compensation), a key factor alongside fears of terminal-value erosion that have driven the massive fall in SaaS valuations. CSU continues to trade at a meaningful discount to its intrinsic value.

SPGI fell roughly 20% from its August 2025 all-time high on fears that AI would disrupt its core business. In reality, its most valuable assets, the S&P 500 index brand, its credit ratings duopoly, and Platts commodities benchmarks, are legal and regulatory standards embedded in trillions of dollars of contracts. The CFO has confirmed that only ~$540 million of revenue faces any meaningful AI substitution risk. If anything, generative AI applications will run on S&P’s proprietary datasets and benchmarks (which drive a 95% of revenue), deepening its ecosystem value.

The near-term catalyst is the spin-off of the Mobility division (CARFAX), which has weighed on SPGI’s multiple. Post-spin, the core business should rerate toward MSCI and Moody’s at 25x+ forward earnings. Our sum-of-parts analysis suggests meaningful upside.

Birkenstock is down over 40% from its peak on FX pressures, U.S. tariff risk, and expectations of a weak consumer.

But the market is ignoring the potential for multiple years of strong growth for this brand. 35% of customers repeat-purchase within two years, up from 20% in 2021. The company sells at over 90% full-price realization (that is, no discounts). Gross margins of ~58% exceed Nike’s and approach levels seen in luxury footwear. FY2025 revenue of €2.1 billion was up 18% in constant currency, with Asia growing 34% (the brand is just starting there). At approximately a 15x PE ratio, you’re paying a stagnated consumer staples multiple for a brand growing at double-digit rates with luxury-level margins.

TIC is the most asymmetric opportunity on our watchlist. Formed by the August 2025 merger of Acuren and NV5, it’s a $2.1 billion revenue platform in testing, inspection, and compliance - a large portion of which is legally mandated by federal regulators. Customers cannot cut these budgets without risking shutdowns and fines. Both legacy businesses experienced only marginal revenue declines during 2008 and COVID. The AI-driven data center boom is directly boosting TIC’s power and utilities segments. The direct revenue from data centers is ~3% of the total, but growing by over 40% year-on-year.

The stock has corrected sharply, due to slower than expected integration and a recent management change. It was sold down to an irrationally cheap level by the market - the recent rebound rectifies a little bit of that.

Executive Chairman Sir Martin Franklin executed an almost identical playbook with APi Group, delivering 25%+ annualized returns over five years. At ~$8 per share, TIC trades well below our bear-case intrinsic value of $16 and our base case of $23 - a stock a $33 billion institutional fund bought via private placement at $12 just months ago.

What connects all of these names? None of them became worse businesses when the Iran War started. Their competitive positions didn’t erode. Their customers didn’t disappear. Their long-term earnings power wasn’t impaired.

What changed was sentiment. And sentiment, unlike fundamentals, tends to correct fast.

The current market environment (with strong volatility) will continue to offer great opportunities to those waiting for the right valuations (as I write this article, Netflix has fallen 10% after its earnings release!).

We at RC will continue to dig into quality companies and wait for the inevitable drawdown. Consider upgrading your subscription to get all our reports as soon as we publish them.

Disclaimer: Rebound Capital’s work is provided for informational purposes only and should not be construed as legal, business, investment, or tax advice. You should always do your own research.