Ranking Backtest

Methodology

The simulated historical performance was generated using the same ranking system and factor weights as the current live rankings published on ClarityVesting.

For each historical period, a ranking was built using only the historical data that was available at that point in time. No future information was used, which avoids look-ahead bias (i.e. using data that would not have been available to an investor at the time).

Companies were selected based on their ranking and then combined into a market-cap-weighted portfolio. As a result, large- and mega-cap stocks have significantly greater weight in the historical portfolio than mid- and small-cap stocks.

A three-month lag was introduced between the ranking date and portfolio construction to further reduce the risk of look-ahead bias resulting from late regulatory filings, reporting delays, or financial restatements.

Each portfolio was then held for five years without rebalancing. At the end of each five-year period, a new ranking was generated and a new portfolio was constructed using the same methodology. This process was repeated throughout the 25-year historical period. The simulated historical return represents the compounded performance of these successive five-year holding periods.

Survivorship bias

The data provider used by ClarityVesting has limited historical coverage of delisted and bankrupt companies prior to 2018. This may introduce survivorship bias, which occurs when historical performance is calculated using only companies that still exist today.

However, the impact of this bias is substantially reduced by the market-cap-weighted nature of the portfolio.

To estimate the potential impact of survivorship bias, we can consider an intentionally pessimistic scenario:

Assume that 1% of the total U.S. equity market capitalization disappears each year through complete bankruptcies, which is likely an overstatement.
Assume that 50% of this disappearing market capitalization is selected in the top 5% of stocks by the ranking system, an intentionally conservative assumption that implies the model performs particularly poorly at identifying financially distressed companies.

Based on this pessimistic scenario, the potential overstatement of the backtested annual returns caused by missing delisted and bankrupt companies has a conservative upper bound of 0.5% per year.

Avoiding overfitting

The ranking system has been designed to place greater emphasis on certain metrics considered more relevant for long-term investors. These adjustments were not made with the objective of maximizing backtest performance.

Optimizing a system solely to maximize backtested returns can lead to overfitting. An overfitted model may perform exceptionally well over a specific historical period while performing poorly in different market environments. ClarityVesting has not optimized the ranking system to maximize backtest results.

Limitations to consider

Backtested returns do not include transaction costs, brokerage fees, taxes, or other investment expenses. These costs vary depending on the investor's broker, country of residence, and individual tax situation.

In many jurisdictions, long-term investing may benefit from more favorable tax treatment than short-term trading.

Backtested performance

The ranking system has been backtested over the last 25 years. Historical portfolios were constructed from different portions of the ranking (top 5%, top 20%, top 30%, and others), combining stocks from all sectors currently covered by ClarityVesting. Stocks were weighted by market capitalization so that larger companies had a greater influence than smaller ones. Each portfolio was held for five years without rebalancing. The analysis uses only information that would have been available to investors at the time.

Ranking ⓘ	Ann. Return ⓘ	Nb. Stocks ⓘ	Max Drawdown ⓘ	Ann. Volatility ⓘ
Top 5%	18.83%	26	58.99%	32.94%
Top 10%	17.79%	51	52.73%	29.74%
Top 20%	15.15%	101	53.89%	27.78%
Top 30%	13.44%	152	52.76%	30.18%
Top 40%	13.00%	202	51.11%	29.96%
Top 50%	12.67%	251	51.08%	29.44%
XLK (Tech ETF)	11.37%	—	62.40%	23.82%
SPY (S&P 500)	9.16%	—	55.19%	18.72%
Bottom 40%	7.83%	202	61.20%	28.10%

The chart below shows the hypothetical growth of a $10,000 investment in the top 5% of ranked stocks compared with an investment in the S&P 500 Index.

Backtested performance is not indicative of future results.

To go further

The backtested returns and rankings are based exclusively on quantitative metrics. They do not consider a company's business model, competitive advantages, risks, growth catalysts, management quality, or other qualitative factors.

The rankings should therefore be viewed as a starting point for research rather than followed blindly. Investors who complement the quantitative rankings with their own research, market knowledge, and understanding of the businesses may identify opportunities that are not fully captured by the ranking system alone.