Skip to content

Commit

Permalink
Add strategy decay
Browse files Browse the repository at this point in the history
  • Loading branch information
miohtama committed Mar 14, 2024
1 parent ef8ba03 commit c622469
Showing 1 changed file with 59 additions and 1 deletion.
60 changes: 59 additions & 1 deletion source/glossary.rst
Original file line number Diff line number Diff line change
Expand Up @@ -2194,7 +2194,7 @@ and algorithmic trading.

Best trading algorithm

The best trading algorithm in the world is one that makes most profit.
The best trading algorithm in the world is one that makes most profit, or :term:`alpha`.

Depending on the the market situation and available :term:`trading pairs <trading pair>`,
the best trading algorithm can vary day by day, or depending whether the markets are in
Expand All @@ -2214,6 +2214,8 @@ and algorithmic trading.

See also

- :term:`Alpha`

- :term:`Profitability`

- :term:`Trading strategy`
Expand Down Expand Up @@ -4544,4 +4546,60 @@ and algorithmic trading.

See :term:`Annual Percentage Yield (APY)`

Alpha

In :term:`quantitative finance`, "alpha" refers to a measure of an investment strategy's performance compared to a benchmark index, after adjusting for risk. It represents the excess return generated by the strategy beyond what would be expected based on its level of risk. Alpha is often used to assess the skill or effectiveness of portfolio managers, traders, or investment strategies in generating returns.

Mathematically, alpha is typically calculated using the Capital Asset Pricing Model (CAPM) or similar models that relate an asset's return to its level of risk. In these models, alpha is the intercept term of the regression equation, representing the excess return not explained by the systematic risk factors.

Positive alpha indicates that the :term:`trading strategy` has outperformed the benchmark, while negative alpha suggests underperformance.

Alpha is a crucial concept in quantitative finance, as it helps investors evaluate the effectiveness of their investment decisions and identify sources of added value beyond market movements.

See also

- :term:`Alpha model`

- :term:`Trading strategy`

- :term:`Quantitative finance`

- :term:`Risk-adjusted return`

- :term:`Alpha generation platform`

- :term:`Profitability`

- :term:`Strategy decay`

Strategy decay

In :term:`quantitative finance`, strategy decay means that a :term:`trading strategy` loses is excessive profit, or :term:`alpha` generation, capabilities over time.

`From QuantStrat <https://www.quantstart.com/articles/annualised-rolling-sharpe-ratio-in-qstrader/>`__:

"Strategy decay is one of the trickiest aspects to manage within the realm of quantitative trading. It involves previously performing strategies that gradually, and sometimes rapidly, lose their performance characteristics and end up becoming unprofitable."

"Quantitative trading strategies almost unilaterally rely on the concept of forecasting and/or statistical mispricing. As more and more trading entities–retail or institutional–implement similar systematic strategies the mispricings give way to price efficiency. The gain derived from such strategies is eroded and then usually falls to the level of transaction costs required to carry it out, making them unprofitable."

"This means that quantitative trading is not a "set and forget" activity. In reality quant traders need to have a portfolio of strategies that are slowly rotated out over time once any arbitrage opportunities begin to erode. Thus constant research is required to continually develop new profitable edges that replace those that have been arbed away".

One common way to measure and visualise strategy decay is visualising rolling :term:`Sharpe` ratio.
If Sharpe is getting worse over time, the strategy is decaying.

See also

- :term:`Sharpe`

- :term:`Trading strategy`

- :term:`Quantitative finance`

- :term:`Alpha`

- :term:`Sharpe`

- :term:`Risk-adjusted return`

- :term:`Profitability`

0 comments on commit c622469

Please sign in to comment.