The Pentagon is funding my next 100-bagger candidate
And Ergodicity?
Hi hunters,
Let’s go through the markets and look at a company that has all the signs of a 100-bagger.
1. Bubble Meter
In our previous edition, we discussed the rising margin debt.
And it keeps on trending upward.
Now I do not want to rehash the entire analysis from last time. We talked about the difference between the SP490 and the SP500, and what all these metrics we track actually mean.
Predicting when a correction or crash happens is impossible. But we should always be aware of what the general market is doing.
2. Stu(o)ck in the funnel
Merlin (MRLN) - the setup in short
What: Boston aerospace-autonomy company builds an aircraft-agnostic autopilot for military + civilian planes
Valuation: $470M market cap on $7.7M trailing revenue (60x sales)
Balance sheet: $183M cash, zero debt; 6–8 quarters of runway
The contract: $105M US Special Operations Command contract to automate the C-130J; GE Aerospace partnership
The bet: First to clear FAA certification for machine learning-in-the-cockpit writes the template everyone copies
The problem: Better-funded rival (Reliable Robotics) may be ahead on the regulator-agreed path
Outcome: Binary - 0 if it loses the race, 100-bagger if it establishes a monopoly
My stance: On the fence, watching from the sidelines
Description: Merlin (Ticker: MRLN) is a Boston-based aerospace autonomy company. It builds Merlin Pilot, an aircraft-agnostic system designed to fly any plane, military or civilian, from takeoff to touchdown. First alongside a pilot, eventually without one. The architecture is a deterministic, rule-based flight core running in parallel with a machine-learning layer. A pure black-box AI cannot be certified for flight, so the machine-learning plans and the deterministic core guarantee the behavior. Merlin went public via SPAC in March 2026. Their anchor contract is a $105M contract with US Special Operations Command to automate the C-130J, plus a partnership with GE Aerospace targeting single-pilot operations. Trailing revenue is about $7.7M. Q1 showed a $90M net loss, mostly non-cash SPAC accounting. Cash is $183M, zero debt.
Type: Pre-revenue defense-tech microcap
Why it’s interesting: A certification race where the customer pays for the R&D, with a binary outcome.
The rundown: At roughly a $470M market cap on under $8M of revenue, you’re paying 60x sales for a company that loses money on purpose. The entire value is future optionality.
Here’s the bet. The flying part of autonomous flight is largely solved. Autopilots have existed for decades. The unsolved problem is certification: the FAA framework requires verifying every output of a flight-critical system, and a machine-learning system has effectively infinite outputs. No accepted compliance method for ML in the cockpit exists yet. Whoever first gets a new autonomy function through FAA certification writes the template that everyone else must follow. That approval, not any funding round, is where the first signs of an actual moat appear.
Merlin’s edge is that defense money pays for the evidence. USSOCOM funds the development; military airworthiness is a lower bar than civil certification; and every assurance artifact generated on the C-130J is partly reusable for the civil case later. Someone else foots the bill. The program passed its Critical Design Review in June, and the stock jumped 32% premarket. For a pre-revenue name, milestones are the earnings reports.
What Merlin does not have is a clear technical lead. The two-layer deterministic-plus-ML architecture is smart, but it’s table stakes. Everyone in the field converged on it. Reliable Robotics claims to have the only FAA-approved certification plan for full automation, to have flown fully uncrewed, and to have raised roughly $294M, compared with Merlin’s $134M in lifetime funding. On the regulator-agreed path, Merlin is behind.
Now the cash position. $183M in the bank. Management guides to $62M in burn for 2026, but if we take the Q1 results, it annualizes closer to $95M. Let’s call it six to eight quarters of runway. The bull case is that defense revenue scales before the market tires of funding the burn. The bear case is that certification takes more time, and they will need to raise more cash.
This is not a compounder, and it is not a value stock. It’s a call option on a regulatory event, partially funded by the Pentagon, with two years on the clock and a better-funded competitor possibly ahead in the queue. The outcome is binary. It could go to 0 because it loses the race. And it could become a 100-bagger if it were to establish a monopoly in autonomous flight.
Current status: On the fence. One thing I did not mention is that insiders are aligned.
Normally I would show you a per-share metrics chart from fiscal.ai, but that doesn’t make a lot of sense here, so here is the insider ownership table:
And that the way the SPAC was structured was unusual for a typical SPAC (meaning some people did not take profits; they went full equity, all in on this company). So I’m watching from the sidelines for the moment. If Merlin has a regulator-agreed compliance path, this setup becomes more bullish (or at least they are actively developing it with the regulator).
3. Best article of the week
What I loved reading this week was an article written by Polymath Investor called:
It explores ergodicity. (There’s a whole book about this topic written by Luca Dellanna)
Ergo… what?
Yeah, I had the same reaction when I first heard about it. This does not comply with my Feynman approach to explaining things.
The basic definition is as follows:
Ergodicity is whether your path through time averages out to the same thing as a snapshot of the whole crowd.
If yes, ergodic.
If not, the crowd’s average is not what you should expect for yourself
A typical example is the following experiment: a coin-flip game.
Heads, you win 50%
Tails, you lose 40%
So if you math this out, 1.5 on every win, 0.6 on every loss, over time, you’ll get 1.05 or 5% on every flip.
But in reality, that’s not what happens.
You don’t get “the average.” You get a sequence. You flip, and flip, and flip, and your money doesn’t add; it multiplies, because each flip acts on whatever you have left. Say you flip heads, then tails. 1.5 times 0.6 is 0.9. You’re down to 90 percent. Do it again: another heads-and-tails, 0.9 times 0.9. Every pair of flips quietly eats a tenth of you. Keep going, and you’ll go broke.
So if 10,000 people at a concert do the coin flip test, the entire crowd will be up 5% per flip. But not everyone in the crowd will be up. Some will be broke.
What can we learn from this:
Survival compounds. The goal is to stay in the game. Maybe you’re down a couple of percent this year. That’s not a problem. Avoid big drawdowns in your portfolio. Those hurt the compounding the most.
May the markets be with you, always!
Kevin