What a 2024 pair trade taught about beta neutral

This is my own research note from 2024, republished here under Goatnews, my personal brand at the time. It's built around Pear Protocol and a live example running October 2023 to May 2024. Every number below is frozen at that write date. Nothing here reflects what happened to FET, ETH, or Pear Protocol after that window closed. The image caption's "Made by Goatnews (Euzinho)" was me signing my own work; Euzinho is Portuguese for "little me."

Long $FET, an AI-sector token, short $ETH, built to stay market neutral.

It played out three ways.

One aimed for a beta of zero. One just matched the dollar amounts. One kept a beta of 0.5, on purpose.

The market didn't hold still for any of them, no matter which target they were built around.

it ran on Pear Protocol, the first onchain platform built specifically for pair trading: buy one token, short another, on margin, in one product.

Long/short can lean into a view instead of chasing pure neutrality, something like a catalyst or a sector rotation, so you're never betting the whole position on one token's direction alone.

The report says plainly it's educational only, not financial advice. I'll say the same about sharing it.

I just want to walk you through the math underneath it, nothing more.

dollar neutral is the easy version: $100k long, $100k short.

It looks balanced on the surface.

Whether it's balanced against the market comes down to beta, how hard each side swings when the market moves.

Say your long side has a beta of 1.4. Your short side has a beta of 0.7.

Split the money evenly between them, and your net beta lands at 0.35.

Half of 1.4, minus half of 0.7.

That's not zero.

A "neutral" book built like that still moves with the market.

To cancel it out, the short side needs to be twice the size of the long side.

1.4 divided by 0.7 is 2. Only then does the market exposure wash out.

all three ran October 2023 to May 2024.

Rebalanced daily on a rolling 60-day beta, capped at 1x, no borrowing on top.

Run Portfolio 1 and the ride stays calm, the swings barely register. Exactly what a beta of zero is built to do.

Run Portfolio 3 and you feel every swing harder, carrying more market exposure by design.

Portfolio 2 is the one that stuck with me.

Its beta wandered instead of holding at zero, landing it with a moderate long bias the whole way through.

Its volatility sat between the other two.

A plain long-only position in FET, no short leg at all, would have earned more than any of the three pair trades.

That was over the same window.

It also carried a lot more volatility, and its drawdowns nearly doubled.

Portfolio 2 called itself dollar neutral, the easy version from the top of this post. Its beta still drifted long the whole run, the same slippage the 1.4/0.7 math above shows once nobody's rebalancing by beta instead of dollars.

If you're running anything long/short right now, it's worth checking whether your legs are sized by beta or just by dollars.

Especially if you're calling it neutral. Tell me what you're seeing.