Marketocracy's Secret Sauce

This is not investment advice.

I am not a registered investment advisor.

Proceed at 100% your own risk.

On second thought, go away. Just buy a market index fund, the fees are low, and you get stock market returns because you're literally buying the stock market.

Still here? Ok. Since this is not investment advice, what is it?

This is really a story about the startup that almost made me rich.

BTW, the stock market is not efficient. I proved this. Please send me Nobel prize... If you invest in the stock market, you don't believe its efficient either.

Once upon a time, there was a gentleman named Ken Kam, who was a really smart guy. He lived in Silicon Valley and he realized that there were no mutual fund managers who actually knew anything about technology. He got some friends to give him some money to invest and started a mutual fund with a partner. His fund invented in Tech, before it was cool. He did stuff like take an Iomega zip drive apart, found out it was mostly Xilinx programmable ASIC parts and when Iomega was telling analysts they were going to get rid of the expensive Xilinx parts, Ken said "Sure, Not" and bought Xilinx.

This made Ken a rebel, because normally, the way Wall Street picks mutual fund managers is the following:

They have to go to the right school

They have to come from the right family.

That wouldn't make a good softball team, so not surprisingly Ken's mutual fund soon had the best 3-year track record in the country. Ken's mutual fund went from millions to billions in assets under management in under six months after the 3 year record was published. Ken was getting 2% of that as a management fee. Pretty sweet.

Unfortunately, like many business partnerships, Ken and his business partner disagreed, so his partner bought him out of the business. Ken was ok with that, because it freed him up to do his new idea "Marketocracy". As someone not from an old Wall Street family, who only went to Stanford, Ken thought the way Wall Street should find mutual fund managers was to give them $1M in fake money and have them run a fund. After 3 years, Ken could find whoever did the best and start a mutual fund based on that.

So we built that back in 2000, a website, you got $1M, we would "trade" the stock on our website, fill orders if the volume allowed, etc. It was a lot of fun to build. But 3 years is a long time to fund a startup out of the buyout from your old company. So the executive team came up with idea "what if we didn't just bet on 1 person, but picked the top 100 people and built a fund of funds?"

That's where I came in. I built a whole system in Perl, (because I'd just read a book on Perl, and Perl is basically just math.) So the system would take a collection of funds and track their performance as a basket, essentially blending the performance. It did pretty well, because we were so broadly invested (3000+ stocks), we were less volatile than the stock market, and we were beating it by 10%/year, which is huge. They made me Director of Fund Analysis. Marketocracy was going to be profitable, and just skip going public entirely. The employees were going to be able to cash out because the company was going to do complicated stock magic to get taxed on capital gains instead of profits, the future was so bright, we had to wear shades...

But then we invaded Iraq. The market got all emotional. We started trailing the market badly, and Ken panicked, and we screwed up the fund.

But that's boring and sad, so I don't want to talk about it.

What I want to talk about today is the method we used that ended up being the best way to pick funds. Ken it called the "batting average".

With the fund-of-funds we were effectively buying a manager for 30 days and then swapping. So the real question for each name was simple: if I buy this fund today and hold it for 30 days, how often does it beat the market over that exact window?

We calculated that percentage across a sliding 30-day window. That single number, the batting average, told us more about consistency than raw trailing returns. We wanted turtles, not hares who sprinted at the last moment.

My Mom died in 2025, a little over a year ago now, and earlier this year the annuity she'd inherited from my dad found me and said "Hey, this is now your annuity". And the Financial Planner who sold my Dad the annuity called me up and said "Your annuity is totally 2000 with the fees, I can do better". Er, you want to give me more money while saving me income taxes? Ok...

But that meant I had to decide if the fund he was suggesting was ok. Nobody gets 2% in management fees anymore. The fund charges .39% more than an index fund, would it be worth it?

So I had grok write me a batting average script, and I tested the fund he suggested, and on a 1 year horizon, yes. On a one-year holding period the fund has beaten the S&P 500 in roughly 62% of overlapping windows, with a positive average excess return. Even restricted to the last five years the one-year win rate is still 64%.

That is not a guarantee. It is simply the same test we used when we were trying to find managers who were more often than not ahead of the market over the horizon that actually mattered to us.

Again, not investment advice. YMMV. Past performance not a guarantee of future performance. Most people should just buy the index. Don't be burdened by what has been.

But if you’re going to pay up for active management, at least demand a batting average that is meaningfully above 50% on the time frame you actually care about.

You can get it from github here: Github Repo