liveBook · Manning

Sharpe ratio

As explained in Applied Reinforcement Learning

A measure of risk-adjusted investment performance, commonly calculated as excess return divided by return volatility.

The Sharpe ratio measures return earned per unit of risk: average return above cash divided by the standard deviation of returns. A higher ratio indicates a smoother path to profit. section 12.8

Use it to compare portfolio strategies on a test window. In one comparison, a fine-tuned-signal RL agent achieved a Sharpe ratio of 8.47, versus 2.62 for the next-highest mode. section 12.8

Table 12.2 Five modes on the test window, averaged over all episodes
Mode
Sharpe
Annualized Return
Annualized Volatility
Maximum Drawdown
Turnover
Final Value
Buy-and-hold SPY
0.77
26.6%
34.4%
−12.33%
1.00
1.0432
Equal-weight
0.91
16.0%
17.5%
−6.34%
1.16
1.0216
RL, no news
1.41
40.9%
29.1%
−9.76%
4.07
1.0571
RL + naive signal
2.52
67.1%
26.6%
−7.06%
6.09
1.0876
RL + base LLM signal
2.62
66.7%
25.5%
−7.07%
6.49
1.0871
RL + fine-tuned LLM signal
8.47
169.5%
20.0%
−2.18%
8.96
1.2251

Compare Sharpe ratios alongside returns, volatility, drawdowns, and turnover across five portfolio modes. section 12.8

How Applied Reinforcement Learning covers it

How deep it goes
Defines the ratio as excess return divided by return volatility and interprets higher values as smoother gains.
Examples
Compares five portfolio modes, including buy-and-hold, equal-weight, and RL agents using different news signals.
What you can do after
Evaluate portfolio strategies by comparing return per unit of risk.
How it connects
Ties risk-adjusted performance to excess return, volatility, and reinforcement learning for portfolio management.

Also explained in

Compare how 3 books explain Sharpe ratio →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Applied Reinforcement Learning. The section links point to the parts of the book it draws on.