[chii]

The stock market is very efficient - aka, the price reflects most, if not all available information at the time.

Therefore, a trading bot would need to trade on information that is not available to other bots, for it to be profitable. This is hard to achieve, and each bot that _do_ achieve it is making the market even more efficient.

[Flankk]

Two investors can look at the same balance sheet and come to opposite conclusions. 90% of actively managed investment funds underperform the market. It's a hard problem, full stop. If you're good at solving hard problems, you could solve something less boring and probably make more.

[westurner]

Managed funds must pay managers; probably out of the returns.

ETFs (Electronically Traded Funds) have no fund management fees; like Class B stock, ETFs typically are not voting shares (which you don't have when you buy a mutual fund anyways).

Algotraders reference e.g. an S&P 500 ETF as the default benchmark for comparing a portfolio's performance.

Quarterly earnings reports are filed as XBRL XML. A value investor might argue that you don't need to rebalance a portfolio until there is new data about technical fundamentals for technical analysis.

The average bear trades on sentiment and comparatively doesn't at all appropriately hedge; this is part of Behavioral economics, the technical reason why some people actually can outperform the market, imho.

[westurner]

If the financial analyst does not have a (possibly piecewise) software function to at least test with backtesting and paper trading, do they even have an objective relative performance statistic? Your notebook or better should also model fees and have a parametrizable initial balance.

Here's the awesome-quant link directory: https://github.com/wilsonfreitas/awesome-quant

[mbrodersen]

Only over the very long term. Short term there is a lot of noise and randomness.