Allocating Capital Across Concurrent Trading Models
Summary
The document raises a portfolio-allocation problem for an algorithmic trader running several stock-market models that may generate signals at the same time. With a finite wallet, committing all capital to one model can prevent acting on another signal and concentrates exposure in a single trade. The question asks whether a mathematical framework can allocate capital across models while preserving diversification.
No answer or allocation method is included, and the expected returns alone would not establish a suitable allocation. The document supplies no estimates of volatility, correlation, trade overlap, drawdown, or capital constraints, all of which would matter in evaluating simultaneous strategies. It therefore serves as a statement of the capital-allocation problem rather than a worked solution or evidence for a particular sizing rule.
Key ideas
- Several trading models can produce signals while sharing a limited pool of capital.
- Allocating the entire wallet to one signal can leave no funds for later opportunities.
- Concentrating capital in one trade increases exposure to model-specific risk.
- A useful allocation method would need more than expected returns, including risk and dependence estimates.
- The document poses the problem but does not provide a mathematical solution.
Tags
Full text
# Multiple models for one single wallet # Multiple models for one single wallet I am an algorithm day trader. I am trying to automate certain patterns on the stock markets. Here is my question. Suppose I have $n$ models $M_1, M_2, M_3, \dots, M_n$ with expected return $E_1, E_2, E_3, \dots, E_n$. Each of those models can be triggered at any time. Suppose I have a wallet of $X$ dollars. How can I handle the models using that wallet? If $M_i$ is triggered, I can't take the whole $X$ dollars for one single trade. If so, then if $M_j$ is triggered, then I can't trade because my wallet has $0$ dollar in it. In addition, as I use my portfolio for only one trade, then I put myself more at risk than if I used more than one (diversification). Is there a mathematical way to deal with that?
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.