> For the complete documentation index, see [llms.txt](https://docs.sov.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.sov.ai/asset-managment/weight-optimization.md).

# Weight Optimization

`Tutorials` are the best documentation — [<mark style="color:blue;">`Weight Optimization Tutorial`</mark>](https://colab.research.google.com/github/sovai-research/sovai-public/blob/main/notebooks/studies/Weight%20Optimization.ipynb)

### Key Features

* Multiple optimization strategies
* Comprehensive performance analysis
* Risk-adjusted return metrics
* Portfolio composition visualization
* Drawdown and contribution analysis
* Correlation and clustering analysis
* Daily weight tracking

### Usage

To use the Weight Optimization module, you first need to prepare your dataset. Here's an example of how to set up and run the optimization:

<pre class="language-python"><code class="lang-python">import sovai as sov

# Authenticate
sov.token_auth(token="your_authentication_token")

# Prepare your data
df_price = sov.data("market/closeadj")
df_mega = df_price.select_stocks("mega").date_range("2000-01-01")
df_returns = df_mega.calculate_returns().dropna(axis=1, how="any")

# Select the most uncorrelated stocks
feature_importance = df_returns.importance()
df_select = df_returns[feature_importance["feature"].head(25)]

# Run weight optimization
<a data-footnote-ref href="#user-content-fn-1">portfolio</a> = df_select.weight_optimization()
</code></pre>

## Overall Portfolio Analysis

### **Sharpe Ratio Distribution**

Shows the distribution of Sharpe ratios across different strategies, helping to understand the consistency of risk-adjusted returns.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>.sharpe_plot
</code></pre>

<figure><img src="/files/AVzlWsDpHS6bj6hn7VJY" alt=""><figcaption></figcaption></figure>

### **Cumulative Returns Plot**

Displays the cumulative returns of all portfolio strategies over time, allowing for easy comparison of overall performance.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>.return_plot
</code></pre>

<figure><img src="/files/zEn5pgkGLPVgN1fkNER1" alt=""><figcaption></figcaption></figure>

### **Overall Composition Plot**

Illustrates the asset allocation of all strategies, allowing for a comparison of how different models allocate capital.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>.composition_plot
</code></pre>

<figure><img src="/files/SyqWH87L4Nl16WcwXLFJ" alt=""><figcaption></figcaption></figure>

### **Best Performing Model**

Identifies the strategy that performed best according to the Sharpe ratio.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>.best_model
</code></pre>

```
'NCO'
```

### **Performance Summary**

Provides a comprehensive summary of key performance metrics for all strategies, including returns, volatility, Sharpe ratio, and more.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>.performance_report
</code></pre>

<figure><img src="/files/8d3arZwf6pFsaAlwfqUi" alt=""><figcaption></figcaption></figure>

## Model-Specific Analysis

For model-specific analysis, replace "model\_name" with the actual model name (e.g., HRP, HERC, NCO, or EQUAL).

### **Cumulative Returns**

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>["model_name"].backtest_plot
</code></pre>

<figure><img src="/files/XeaxKWzL0lPfaC3Xin9q" alt=""><figcaption></figcaption></figure>

Displays the cumulative returns of the specific model over the backtesting period.

### **Backtest Report**

Detailed performance statistics from the backtesting period for the specific model.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>["model_name"].backtest_report
</code></pre>

<figure><img src="/files/BSQSFqASZHq9u0KmPVhX" alt=""><figcaption></figcaption></figure>

### **Rolling Sharpe Ratio**

Visualizes how the Sharpe ratio of the model changes over time, indicating consistency of performance.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>["model_name"].sharpe_rolling_plot
</code></pre>

<figure><img src="/files/5EH4ZFt4PSd8ceuAHC1A" alt=""><figcaption></figcaption></figure>

### **Model Composition**

Illustrates the asset allocation for the specific model.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>["model_name"].composition_plot
</code></pre>

<figure><img src="/files/Bcs8tyHccdOWFQ3yBiEg" alt=""><figcaption></figcaption></figure>

### **Drawdown Contribution**

Shows which assets contribute most to the portfolio's drawdowns, helping identify risk sources.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>["model_name"].drawdown_contribution_plot
</code></pre>

<figure><img src="/files/DzqhJmiaKdTzKiiLxe1r" alt=""><figcaption></figcaption></figure>

### **Sharpe Ratio Contribution**

Indicates which assets contribute most to the portfolio's Sharpe ratio, highlighting return drivers.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>["model_name"].sharpe_contribution_plot
</code></pre>

<figure><img src="/files/744RwC5gNufmu2FzJmvD" alt=""><figcaption></figcaption></figure>

### **Correlation Heatmap**

Displays the correlation structure of assets used in the model (not available for EQUAL).

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>["model_name"].heatmap_plot
</code></pre>

<figure><img src="/files/f11sSY84UcvE8nWPFYJ7" alt=""><figcaption></figcaption></figure>

### **Clustering Dendrogram**

Visualizes the hierarchical clustering of assets used in the model (not available for EQUAL).

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>["model_name"].cluster_plot
</code></pre>

<figure><img src="/files/i3NI4U6ASrzetKFVBcmD" alt=""><figcaption></figcaption></figure>

### **Current Recommended Allocation**

Provides the model's most recent recommended asset allocation.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>["model_name"].recommended_allocation
</code></pre>

<figure><img src="/files/CvguDYlaJsLBB6vMMD4x" alt=""><figcaption></figcaption></figure>

### **Sharpe Ratio Distribution**

Shows the distribution of Sharpe ratio helping to understand the consistency of risk-adjusted returns.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>["model_name"].recommended_allocation
</code></pre>

<figure><img src="/files/SOp32Hp8xg5Dd8O7XtEK" alt=""><figcaption></figcaption></figure>

### **Daily Weights**

Shows how the model's asset allocation changes day-by-day over the backtesting period.

<pre class="language-python"><code class="lang-python"><a data-footnote-ref href="#user-content-fn-1">portfolio</a>["model_name"].daily_weights
</code></pre>

<figure><img src="/files/wSD4hmJb2FgsmZJcaAFV" alt=""><figcaption></figcaption></figure>

[^1]: class module