How I Size Lottery-Ticket Stocks

These are personal research notes, not investment advice. Please read the investment disclaimer before acting on anything here.

For background on how these notes are created, see the About page. I am not a stock analyst. I am sharing these notes partly as memory joggers for myself and partly because I have learnt a lot from other private investors sharing theirs.

This approach is influenced by the Taleb barbell. In finance, the basic idea is to keep fragile, high-upside bets small enough that failure is survivable, while preserving enough exposure for rare outliers to matter. I am not following the pure version where most capital sits in risk-free assets; my SIPP portfolio is deliberately growth-focused. But the barbell idea is part of how I size the highest-risk/outlier sleeve.

That means I need a plan before signals arrive. Most of these names start as raw ideas from Jason Needham’s Trading Bases membership and his watch lists / Leaderboard, which screen for growth, value, beta, market cap, primary trend, and secondary trend. Before I buy, pass, or replace an existing holding, I review the business case and risks for myself and check them against my remaining risk budget and the major index trend models that shape how much exposure I want on.

My approach to bet size and risk management depends on whether the company is pre-revenue or already has revenue and profits. The technical setup still matters in both cases, but pre-revenue resource names are speculative tickets; operating businesses can justify a different sizing method.

Around this time, three raw signals arrived together: EST, HREE, and THX. They show the practical difference: two names where the whole position is the bet, and one where the initial risk is set from entry to stop loss.

Two Tickets

EST and HREE both belong in the lottery-ticket sleeve, but for different reasons.

EST is a Kazakhstan explorer where the case depends on copper demand, geology, partner support, and funding itself long enough for either Verkhuba, Rulikha, or the gold package to matter to the equity. There was no turnover from 2021-2025, year-end cash was only £442k, and shares outstanding rose from 69.5m to 475.2m. The separate EST note carries the asset details and the ShareScope charts; the comparison point here is simpler: I do not treat this as a business-quality position. It is a small paid option on exploration progress and partner-funded development.

HREE has the more obvious strategic theme: rare-earth exposure outside China is scarce. Ampasindava is a Madagascar ionic clay rare-earths project with a pre-feasibility study showing US$249.6m post-tax net present value at a 10% discount rate (NPV10) against US$142m of pre-production capex. The risk/reward is visible against a market cap around £19m, but the HREE note is the reminder that those project numbers sit above a tiny, loss-making listed company. The upside is strategic-financing asymmetry; the failure mode is dilution or stagnation while the project still looks good on paper.

For both names, the bet size is about 0.33% of the portfolio, or 33 basis points. That is the whole position. It is not a risk-to-stop calculation. In this part of the portfolio, I assume many names will either go nowhere or fail outright. The bet size has to be small enough for that assumption to be tolerable.

One Profitable Business

THX is different. Thor Explorations is a West African gold producer with one producing mine, Segilola in Nigeria, and one important development project, Douta in Senegal.

ShareScope showed 2025 production of 91,910 ounces, revenue of US$325.5m, free cash flow of US$180.7m, cash of US$137.8m, and total borrowing down to US$2.6m. The THX note carries the charts and the valuation table; the comparison point here is that this is already an operating business.

That does not make it low risk. One producing asset still carries too much of the story, and Douta introduces its own development and financing risks. But the sizing method changes. For THX, initial risk is 50bps from entry to stop, and position size is reverse-engineered from that. The risk is still explicit, but it is not the same as buying a tiny wipeout-prone ticket where the whole gross position is the risk.

The Lottery-Ticket Sleeve

Lottery tickets can be up to 5% of the portfolio in total, with each starting bet sized at about 0.33%. In practice that means no more than roughly 15 names if the sleeve is full.

The assumed outcome distribution is harsh. I expect most of these names to disappoint. A working assumption is roughly 80% losers and 20% winners, with some losers simply drifting sideways or down and others going to zero. The point of the sleeve is that a few winners, including the occasional very large winner, have to carry the whole basket.

The second rule is not owning too many versions of the same bet. I do not want all 15 names to be tiny miners, battery-metal explorers, rare-earth developers, or pre-revenue biotech names. As a norm, I want no more than two names of the same broad risk type in the lottery-ticket sleeve. If a better one comes along, the decision is replacement or abstention, not automatic accumulation.

The third rule is that I rarely add to these positions. I may add only if the business starts proving itself, the market confirms it, and the chart gives a defined add point. Strength can earn more capital. Weakness does not.

The comparison to keep in view is simple: is this a speculative ticket or a real operating business, and what does that mean for position size and portfolio slots?