Actionable insights — Comstock, Liberty Stream & Merlin

3:35 1. Barbell the book — compounders vs call-options, sized by rule

1. Split the portfolio into two buckets: a slug of cheap, cash-generating, growing compounders, and a bucket of high-variance bets treated like call options (warrants, LEAPs, or pre-revenue early-stage names).
2. Enter the option-like names small (2–3%) and let winners run; manage risk through position sizing, not stop-losses.
3. When the option bucket balloons (a few names go up multiples), don't add — "rob from Peter to pay Paul," trimming to keep neither bucket out of whack.
4. Be "ruthless about never adding to" an options position — the going-to-zero probability is real and ever-present.

• Any single high-variance name drifting to a double-digit weight — the signal to trim, not add.

11:22 2. Underwrite the multi-year DCF — isolate the one variable that matters

1. Judge risk-reward against the outlook one-to-three years out, not today; a company is worth its multi-year future cash flow, not the next 12 months.
2. Separate structural from cyclical: identify the one multi-year swing variable and largely ignore near-term noise around it.
3. Weight your attention to risks that "permeate beyond a 12-month horizon"; near-term uncertainty is usually already in the price.

• Names where the bear case is purely near-term (timing, one quarter) while the multi-year economics are intact — that's a mispriced clock, not a broken thesis.

13:45 3. Avoid reflexive cash-burners — map the funding ladder

1. Screen out explorers and pre-cash-flow names where the next step must be funded by raising capital — a missed milestone makes dilution reflexive (low price → bigger raise → more dilution).
2. If you do own one, map every funding step and the "timer" on it; a screw-up at any step compounds.
3. Prefer businesses already generating cash or with assets they can sell to self-fund the next step.

• A depressed market cap colliding with a near-term capital need — the setup that "hurts even more" if it goes wrong.

30:37 4. Asset-coverage — get the operating business (and its options) for free

1. Value the saleable / independent assets on their own (here: the land, the mining claims) against the whole enterprise value.
2. If the cash those assets could realize approaches or exceeds the entire EV, you're buying the operating business — plus any call options — for nothing.
3. Then the only thing you must underwrite is that the "free" business clears a low bar (doesn't burn cash / reaches profitability), with the scale of the upside as the debate.

• A monetizable hard asset (real estate, claims, a stake) whose conservative value brackets the whole EV — the margin of safety under an option-rich story.

36:40 5. The vertical-integration test — does it capture the refining step?

1. For a recycling/processing business, ask where the economics leak: shipping intermediate product to a third-party refiner can hand away half the margin.
2. Reward in-house refining/extraction — it lets the tipping fee be cut hard while protecting unit economics, and forces any competitor to solve both the processing plant and the refining (a double moat).
3. Distinguish "does it work at all" (de-risked once a pilot runs) from "how profitable does it scale" — the latter is where the multi-bagger lives.

• Processing plays that quietly give up the refining margin — and the ones moving a pilot to demonstration scale to keep it.

43:25 6. The financing-unlock catalyst — the offtake re-rates the capital stack

1. For a pre-scale producer, recognize the re-rating trigger is rarely the product — it's securing an offtake that unlocks cheaper capital.
2. An offtake enables project-level debt (instead of diluting at the parent), pref financing that can convert at a later up-listing, and a lower cost of capital from a US listing / re-domicile.
3. Grade the offtake's shape: a multi-year deal with floor pricing (uranium-style) is worth far more than a rolling three-month spot arrangement.
4. Treat sizable government (e.g. DOE) funding as an upside option you don't rely on.

• The offtake announcement and its terms (floor vs spot, duration); a US listing / re-domicile and any DOE award as cost-of-capital step-changes.

53:31 7. The listed-VC / de-SPAC screen — find the good cohort, read the signals

1. Start from the base rate: 90%+ of de-SPACs are terrible — hunt the small cohort that are (or could be) genuinely good businesses.
2. Demand a big-TAM, disruptive first-mover with a structural edge (platform-agnostic, not a me-too).
3. Read the insider signal: zero insiders selling — better, insiders/private owners buying as they move into the listed sphere (the opposite of the usual SPAC cash-out).
4. Take incumbents partnering instead of self-building as third-party validation and de-risking.
5. Then size it like venture: small, expect dilution / drawdowns / delays, and be prepared to hold 3–5 years.

• De-SPACs with founder-led ownership, insider buying and incumbent partnerships — and the "hair" (convertible prefs, re-strikes) that defines the real downside.

59:14 8. The dual-use adoption sequence — fund on defense, dream on commercial

1. For dual-use tech, recognize the military steps up first — bankrolling, partnering and de-risking the product.
2. Underwrite the base case on the funded defense pathway that's in front of you; treat the larger, slower commercial market as the "lottery ticket," not the base case.
3. Favor AI that delivers visible value behind real barriers — AI fused with physical product and a long regulatory lead time is hard to replicate.

• Dual-use names where the defense contract funds the build and the commercial TAM is the free option.

1:01:40 9. Probability-weighted bear/base/bull — with an honest $0 bear

1. Write explicit bear / base / bull scenarios and assign probabilities, not a single price target.
2. For near-binary names, recognize that small shifts in those probabilities move the equity a lot — so size accordingly.
3. Always define the genuine zero: read the capital structure for the "hair" (convertible prefs that sit over the equity, re-strikes that lift the share count) that makes a real $0 bear.

• Convertible prefs, re-strike triggers and ADV-linked dilution clauses — the difference between a "high bear" and a real zero.

7:55 10. Scale against the froth — take risk off when you feel most bullish

1. On really frothy days, "scale away from what you're feeling" — if you feel super bullish, that's when to take some risk off.
2. For options specifically, scale out after a large run while they're still out-of-the-money, because the volatility premium is rich and reverses fast.
3. Keep it personal: match the cut to your own risk tolerance and "sleeping level."

• A parabolic, news-driven move in an out-of-the-money option — the moment to scale out, not add.