Previous articles:
Why “buy the top” strategy doesn't work
😆Imagine an investor walks into a financial advisor’s office: “I want to diversify my portfolio!” The advisor nods: “Great. Let’s buy BTC, ETH, and… a little more BTC through an ETF fund.”
This is called the illusion of diversification, where you technically have many different assets, but they all drop in value because of the same news headline.
We will look at three index construction strategies using the example of the Envelop platform’s L1 index, which includes six underlying blockchains: $BTC, $ETH, $BNB, $TRX, $NEAR, and $ZEC. And we’ll use backtesting (that is, “historical testing”) to see how each strategy would have performed from 2023 to 2025 with an initial investment of $100.
Three Ways to Allocate Weights in a Portfolio
- Equal Weight (EW), Equal Shares. Each of the assets receives exactly the same share. Simple, transparent, and fair. It’s like sharing a pizza with friends: everyone gets one slice, and no one feels left out. In May 2026 it's 1/6 ≈ 16.7% of each token in Envelop index.
- Market-Cap Weight (MCap) by market capitalization. Market capitalization (market cap) = token price × number of tokens in circulation. The “more expensive” a token is in this sense, the larger its share in the index. At the start of 2023, BTC accounted for 62.4%, BNB for just 7.4%, and NEAR for 0.3%.
- Correlation-Adjusted Weight (Corr-Adj), adjusted for correlation. Here we use math: assets with high volatility and high correlation with the rest are given a lower weight. The idea is to let “volatile” tokens take up less space in the portfolio.
What Is Correlation and Why Is It Important
Correlation is a measure of how closely two assets move in tandem. A value of +1.0 means “always together,” 0 means “independent,” and -1.0 means “always in opposite directions.”
For our L1 index, the average pairwise correlations look like this: BTC and ETH move together with a correlation of 0.82, which is very high.
Correlation Matrix (2023–2025)
Nore! The asset correlations may vary across different time frames
How correlation-adjusted weights work in a portfolio
The idea is simple: if an asset is risky (high volatility) and moves in tandem with others (high correlation), why hold a large position in it? It doesn’t provide any real protection for the portfolio.
Wi ∝ 1 / (Σi × (1 + Pi)) Σi = the annual volatility of asset i (how “risky” it is) Pi= the average correlation with other assets (how much it “moves in tandem” with them) The smaller Σ and P are, the GREATER the weight
Results for the L1 index: BTC receives a slightly higher share, 21% instead of 16.7%, because, even with low Σ and “systemic” correlation, it remains the most “mature” asset. NEAR receives a smaller share (13.1%) because, with Σ= 100% per year and a correlation of 0.66, it is too unpredictable, while failing to provide enough diversity.
💡 Key Point: ZEC volatility is high (90%), but its average correlation is the lowest in the index (0.63). This means that while it’s “volatile,” it behaves more independently. That is precisely why Corr-Adj allocates 14.8% to it, almost as much as in the equally weighted index. The MCap index, however, allocates only 0.1% to ZEC, and thus missed its +809% rally in 2025.
3-Year Backtest: Which Strategy Made More Money?
Backtesting is like a “time machine” for investing. We take historical asset prices and ask: “What if I had invested $100 in each strategy at the beginning of 2023?”
Sharpe Ratio: Return Per Unit of Risk
The Sharpe ratio is perhaps the most important metric when comparing strategies. It answers the question: “How much did I earn above the risk-free rate for every percentage point of risk taken?”
The formula is simple:
Sharpe Ratio = (Rportfolio − Rfree) /Σportfolio Rportfolio = portfolio return for the year Rfree = risk-free rate (≈ 4%, US Treasury yield) Σportfolio = volatility (standard deviation) of returns
Corr-Adj showed a Sharpe ratio of 9.50. This reflects not only high returns but also consistent growth: +89%, +100%, +110% year over year with almost no sharp declines. MCap, with a Sharpe ratio of 1.11, pales in comparison: its 2025 performance was negative due to its concentration in BTC (which barely rose) and ETH (which fell by 14%).
⚠️ Important! High Sharpe ratios in a three-year backtest do not guarantee future results. Three years is too short a sample size to draw statistically significant conclusions. Truly “good” strategies are tested using 10+ years of data and are verified for resilience against “black swan events.”
Math Over Intuition
Correlation-adjusted weights don't guarantee maximum returns, in some years, they underperform an equally weighted approach. However, they deliver more stable growth and a better risk-return ratio (Sharpe ratio of 9.5 versus 5.76).
The main enemy of a beginner investor is not crypto volatility, but the illusion of diversification in a market-weighted portfolio where 90% = BTC + ETH. A smart index with the right weights is not magic, but the application of Modern Portfolio Theory to the crypto market.
How it works in Envelop
To make it easier for new users to get started and to simplify the understanding of how the Envelop index is built, we decided to use asset-weighted indices in the default presets.
However, you can always create indices with any weightings and any assets.
The Envelop platform offers maximum flexibility.
