I'm putting $50,000 into the smallest stocks on the market

I have $50,000 set aside, a research system that has made exactly zero dollars, and no track record. I’m going to put that money into U.S. microcap stocks and write down everything that happens.

Not the highlights. Everything. The tickers, the share counts, the entry and exit prices, the fees, the dollar result. If I lose $14,000 in a month, you’ll see a chart with a line going down and a post explaining how I managed it.

I write software for a living. I am not a trader. My qualifications for this project are a stubborn belief that most of the stock market is too efficient for someone like me to beat, plus a suspicion that the smallest corner of it might not be. This blog is how I find out which of those two beliefs is wrong.

The stake$50,000Fixed. Additions get disclosed.

Share of my investments~5%The other 95% is in index funds

Deployed so far$0Not until the backtest earns it

Trades0Every one will be logged here

Why the smallest stocks

A microcap is a public company worth roughly $50 million to $300 million.¹ That’s tiny by stock market standards. The average company in the largest tenth of the U.S. market is worth about $334 billion, more than 2,000 times the average company in the smallest tenth.²

Here’s my bet, in one paragraph. Big money can’t play in microcaps. A fund managing $10 billion can’t build a meaningful position in a $150 million company without owning a huge slice of it, and it can’t buy the shares without sending the price through the roof. An SEC staff study found that most stocks under $100 million trade less than $25,000 worth of shares on a typical day.³ So the professionals mostly stay away, and so do the analysts who work for them. That leaves thousands of real companies with real businesses that almost nobody with serious resources is paying attention to.

With $50,000 I’m too small for any of that to matter. That’s the one structural advantage a retail investor has, and I want to know whether it’s worth anything. The next post runs the numbers on how lopsided this gets.

There’s a catch, which I’ll get to: microcaps are also where a lot of people lose a lot of money. The lack of attention cuts both ways.

Why only 5%

This money is about 5% of what I have invested. The rest sits in boring index funds that I don’t plan to touch or write about.

The 5% cap is the most important risk rule in this whole project, and it’s the only one I’m completely sure is right. Microcaps can and do go to zero. If I’m wrong about everything, I want the damage to be a bad year, not a bad decade.

Stake goes to zero

−5.0%

Stake loses half

−2.5%

Stake goes nowhere

0.0%

Stake gains 50%

+2.5%

Stake doubles

+5.0%

−6%−4%−2%0%2%4%6%

Simple arithmetic: a 5% sleeve moves the whole by 5% of its own return.

Show the data

	Change in total portfolio
Stake goes to zero	−5.0%
Stake loses half	−2.5%
Stake goes nowhere	0.0%
Stake gains 50%	+2.5%
Stake doubles	+5.0%

You can read that chart two ways. The pessimist’s version: even if I double my money, my overall portfolio moves by 5%. The optimist’s version: the worst realistic outcome is survivable, which means I can make decisions without panicking. I’ll take the second one. Panic is expensive.

The rules I’m publishing under

Writing about stocks this small comes with a real responsibility. Microcaps are thinly traded. If a few hundred people read about a $40 million company and buy it the same morning, the price moves. That’s exactly how pump-and-dump schemes work, and the SEC has a whole page warning people about them.⁴

So I’m publishing under rules I can’t quietly bend:

1. Every post discloses my position in any ticker it names, as of the publish date. You’ll find the box at the bottom of every post, including this one, which names no tickers.
2. I won’t trade a ticker for 48 hours after a post that names it. I’m not going to sell into attention I created.
3. Losers get the same font size as winners. The Ledger shows every closed trade, net of fees.
4. No editing history. Mistakes get corrected, and the correction gets a date.
5. None of this is advice. I’m one person running an experiment in public. Treat it like a lab notebook.

What I’ll share and what I won’t

I’ll share the build, the reasoning, the trades, the results, and the lessons, especially the expensive ones.

I won’t share the recipe: what exactly the system looks for, how it scores what it finds, and the rules for getting in and out. Part of that is self-interest, since it might be the only edge I have. But in stocks this thin, any edge that a few thousand readers copy stops being an edge for everyone, readers included. I’d rather show you how I think than hand you a signal that breaks the moment you use it.

Where things stand

I started building the research system on September 20. It’s a few weeks old, I’ve worked on it nearly every day, and it has not placed a trade. That’s on purpose. Nothing goes into the market until a historical test that I trust says it should, and building a historical test I actually trust is most of the work.

1. Level 1
   Collect BuiltPull public company data from free sources, politely and reliably.
2. Level 2
   Remember BuiltStore everything exactly as it looked on the day it became public, so the past can be replayed honestly.
3. Level 3
   Analyze BuiltTurn raw documents into structured facts a computer can compare.
4. Level 4
   Backtest In progressAsk whether any of it would have made money after real-world costs.
5. Level 5
   Paper trade Not yetRun it live with fake money and see if reality matches the backtest.
6. Level 6
   Real money Not yetThe $50,000 goes in. The Ledger starts moving.

There’s a real chance the backtest says no. If it does, I’ll write that post too. “Engineer spends months building a system and learns the market is efficient” is a less exciting headline, but it’s still useful, and it’s still cheaper than learning it with real money.

What happens next

Next up: why microcaps at all, with the numbers on how small, ignored and expensive to trade this corner of the market really is. After that comes how I’m trying not to fool myself, because the easiest way to lose money in microcaps is to believe a backtest that’s lying to you.

If you want the whole picture, start here. If you only want to know whether I’m up or down, bookmark the Ledger. Right now it says $50,000 and zero trades. That will change.

Footnotes

1. The SEC’s investor education site describes microcaps as companies “with a market capitalization of less than $250 or $300 million,” and calls the ones under $50 million nanocaps. Investor.gov glossary. ↩

2. Kenneth French Data Library, “Portfolios Formed on Size” (CRSP data through August 2026), average firm size in the top and bottom deciles: $334.2 billion vs. $139.9 million. Data library. ↩

3. Charles Collver, SEC Division of Trading and Markets, “A characterization of market quality for small capitalization US equities” (2014). PDF. ↩

4. “Whoever is promoting the stock may stand to profit at your expense from pumping up the stock price and then selling shares.” SEC, Investor.gov: Microcap Fraud. ↩