Monthly Research Letter | January 2026

The risk environment in early 2026 is defined by a significant divergence between the top-tier technology leaders and the broader market. Our analysis, which filters out stocks with long-term bearish trends (5 and 3 years), current rolling tops, double tops, and high drawdown risks, reveals a highly concentrated universe of suitable assets. This highlights a market undergoing a period of selective strength.

For a London-based steward, an additional layer of complexity is the FX adjustment. The relative strength of the Pound has meant that the pure US Dollar returns have been slightly muted upon conversion. The variance of individual stock returns, once adjusted for currency movements, means the Kelly Criterion heavily penalises single stocks in favour of concentrated index trackers, despite the rigorous filtering process.

Why the Core Framework Remains Appropriate

The core framework is anchored in probabilistic sizing rather than narrative chasing. By excluding index trackers and focusing solely on individual components, we isolate the specific idiosyncratic "edge" of the highest-performing assets.

Excluded Assets (January 2026 Filtering):

• Bearish Trends (1Y/3Y/5Y): Intel (INTC) and Tesla (TSLA) remain excluded due to persistent failure to maintain long-term log-plot support.

• Rolling/Double Tops: Apple (AAPL) and Microsoft (MSFT) have been removed as they test significant multi-month resistance levels with fading momentum signals.

• High Drawdown Risk: Palantir (PLTR) and Super Micro (SMCI) are excluded due to extreme valuation-driven volatility despite high growth ratings. 

Uncapped Analysis: Theoretical Maximum Growth

This model illustrates the theoretical optimal weightings based on the continuous Kelly Criterion (𝑓*=𝜇/𝜎2). In a tracker-free environment, the formula concentrates heavily on high-momentum AI-infrastructure and cloud leaders that have maintained "green light" status. 

• 1. NASDAQ Composite: 54%

  2. SNP500: 38%

  3. NVIDIA: 8%

Note: All filtered stocks were removed, including Apple, Microsoft (rolling tops), Palantir (high drawdown risk), and Intel (bearish trend).

The Capped Model (Risk-Balanced Baseline)

This model illustrates how a 25% cap can be used as a discipline anchor to mitigate concentration risk. By forcing allocation out of the highly favoured trackers, capital is spread across the remaining "next best" stocks, providing a more stable capital stewardship baseline.

1. SNP500: 25%

2. NASDAQ Composite: 25%

3. BSE Sensex: 21%

4. FTSE100: 15%

5. Nikkei 225: 8%

6. TSX: 6%

Performance Analysis (Feb 2021 – Jan 2026)

Illustrating the 25%-capped model: If a steward had deployed £1,000 on the last day of February 2021 and rebalanced annually, the portfolio would have grown to approximately £2,134.40 by January 13, 2026. This represents a total 5-year return of +113.4%. The uncapped model theoretically yielded higher total growth but did so with a maximum drawdown exceeding 45% in 2022, a volatility level that frequently challenges investor discipline.

What We Are Deliberately Not Reacting To

We are deliberately not reacting to the recent "rolling top" patterns observed in Apple and Microsoft. The filtering process mandates that we remain anchored in assets with strong technical signals. We acknowledge that the discipline to ignore short-term narratives and stick to the framework prevents reaction and ensures long-term capital survival.

Why Position Sizing and Risk Balance Matter More Than Ideas

The long-term survival of capital depends less on the "quality" of an investment idea and more on the discipline of its sizing. Probabilistic thinking requires us to accept that any market thesis—no matter how compelling—can be wrong. Risk balance ensures that no single failure results in terminal loss. By sizing positions according to their variance and edge, the steward moves from "picking winners" to "managing outcomes," providing the only stable baseline for survival under uncertainty.

For an investor based in London, the 25% cap is particularly optimal due to currency concentration risk. Since the US and Indian trackers are typically denominated in or correlated with the USD, an uncapped model leaves you over 75% exposed to USD/GBP fluctuations. A 25% cap forces allocation back into domestic Sterling (FTSE 100) and other currency blocks (Euro/Yen), providing a natural hedge against Sterling strength.

Expanded tiers provide additional frameworks and a deeper comparative context regarding the evolution of these global weightings across different liquidity cycles. This Tier-1 letter remains your reference frame.