Skip to main content

​
📊 Bullishness & Rebalancing

The core of Backerville’s approach is our proprietary collective intelligence system that aggregates insights from startup employees.
  • Quarterly Voting: Villagers vote on:
    • Companies they most admire in their sector
    • How bullish they are about their own company (1-100 scale)
  • Admiration Index: These votes form a data-driven ranking of companies
  • Rebalancing: Villages periodically adjust their portfolio based on collective wisdom

​
🧠 The PageRank-Inspired Algorithm

Our rebalancing algorithm combines two critical forms of information to determine company weights:
  1. Insider Bullishness: Only employees with 12+ months tenure can rate their own company
  2. Peer Admiration: Employees identify up to seven (7) companies they most admire in their sector
  3. Network Effect: Companies admired by other highly-rated companies get higher weights
  4. Convergence: The final scoring represents balanced collective intelligence
Our approach is inspired by the PageRank algorithm that powers search engine rankings, but specifically adapted to the unique characteristics of private company valuation and employee insights.

​
Why This Works

This dual approach creates a ranking that’s resistant to manipulation and captures both quantitative assessment and qualitative respect across the startup ecosystem:
  1. Self-Assessment: Employees have unique insight into their own company’s prospects
  2. Peer Recognition: Companies admired by their peers, especially those that are themselves highly regarded, deserve greater weighting

​
🧩 Key Algorithm Components

Data Collection

How we gather and validate employee votes and admiration

Algorithm Details

The complete mathematical model and iteration process

Edge Cases & Refinements

How we handle special situations and prevent manipulation

Rebalancing Implementation

How algorithm outputs translate to portfolio adjustments

​
🔄 From Votes to Portfolio Weights

Our system follows a straightforward process from employee inputs to final portfolio weights:
  1. Data Collection: Employees provide bullishness ratings and admiration votes
  2. Pre-Processing: We apply weights to employee votes based on tenure and seniority
  3. Algorithm Execution: Our iterative algorithm computes optimal company weights
  4. Normalization: Results are normalized to create portfolio allocation percentages
  5. Rebalancing Actions: Villages adjust their holdings based on the new target weights
By combining the wisdom of startup employees across companies, we create a more robust assessment of company potential than any individual employee or company could achieve alone.
The core algorithm iteratively calculates a score for each company that balances its employees’ bullishness with how much it’s admired by other companies:Si(t+1)=(1−δ)⋅Bi100+δ∑j=1n(Sj(t)⋅wji)S_i^{(t+1)} = (1 - \delta) \cdot \frac{B_i}{100} + \delta \sum_{j=1}^n \left( S_j^{(t)} \cdot w_{ji} \right)Where:
  • SiS_i is the score for company ii
  • BiB_i is the bullishness score from employees at company ii
  • wjiw_{ji} represents how much company jj admires company ii
  • δ\delta is a balancing factor (typically 0.85)
For full mathematical details, see the Algorithm Details page.