# Why does my PPO network always collapse to buy and hold? — retained record

> Retained reddit record from reddit:algotrading.

Canonical: https://highsignal.app/data/records/315747ca011dedaa

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:algotrading)  published 2026-10-23 18:09 ingested 2026-10-24 11:02

[original source ↗](https://www.reddit.com/r/algotrading/comments/1xdlhqz/why_does_my_ppo_network_always_collapse_to_buy/)

## retained text



```
I have a 10 year dataset across a range of companies. Unfortunately it's only 1m candles as I'm just doing initial exploration at the moment.

My training loops always seem to collapse to buy and hold across the 9 year test data set. 

Why and what are typical ways to shake it up to prevent this one position learning? Or, do I stop early and fly on one wing and a prayer?

Should I also experiment with A3C?
```



document fetched 2026-10-24 10:45 · canonical https://reddit.com/r/algotrading/comments/1xdlhqz/why_does_my_ppo_network_always_collapse_to_buy
