Weighting & Risk Methods

MarketHeist Portfolio can do more than test a fixed allocation — it can show you where your risk actually comes from, how your holdings cluster together, and re-weight your portfolio by several professional risk-based methods with one click. Everything here is computed from real historical returns in your browser. None of it is investment advice.

Where your risk comes from

Your capital allocation (how your dollars are split) is not the same as your risk allocation (how your portfolio's ups and downs are split). A volatile holding contributes far more risk than its dollar weight suggests.

The "Where your risk comes from" panel shows both, side by side. The classic example: a 60% stocks / 40% bonds portfolio is only 60% invested in stocks, but stocks typically drive ~90% of its risk — because stocks swing much harder than bonds. A low- or negatively-correlated holding (a hedge, gold, long-volatility) can contribute less risk than its weight, and occasionally a negative amount (it reduces total risk).

How it's computed: each holding's share of total portfolio variance, from its historical volatility and its correlation with the rest of the book. The shares add up to 100%.

How your holdings cluster

The dendrogram ("How your holdings cluster") groups your holdings by how they move together. Branches that join closer to the left are more correlated; holdings that only join near the right barely move together. Two S&P 500 funds join almost immediately; stocks and long-term Treasuries join much later.

This matters because assets that move together don't diversify each other — three tech-heavy funds are closer to one big bet than three independent ones. The clustering is also what the Hierarchical weighting methods below use to allocate risk fairly across groups, not just individual tickers.

Weighting methods

The "⚖ Balance risk by…" menu (on an editable portfolio) sets your weights by a chosen method. All are long-only (no shorting) and sum to 100%. You can always edit the result by hand afterward.

Equal weight

Every holding gets the same weight (1/N). Simple and transparent, but it ignores that some holdings are far riskier than others — so your risk usually ends up concentrated in the most volatile positions.

Inverse volatility

Each holding is weighted inversely to its own volatility — calmer assets get more, jumpier assets get less. This balances risk better than equal weight, but it looks at each holding in isolation and ignores how they move together.

Equal Risk Contribution (ERC)

Also called risk parity. Weights are set so that every holding contributes the same amount of risk to the portfolio. This accounts for both each holding's volatility and its correlations. It's the most balanced from a pure risk standpoint — the "effective number of bets" (see below) is maximized.

Note that ERC often puts a large capital weight on very calm assets (like short-term Treasuries) to bring their small risk up to parity with everything else — so a high dollar weight there is expected, not a bug.

Hierarchical Risk Parity (HRP)

HRP (introduced by Marcos López de Prado) first clusters your holdings by correlation — the dendrogram above — then allocates risk down that tree. Instead of treating all holdings as independent, it splits risk between groups first, so a cluster of similar assets can't quietly dominate. HRP is known for being stable and often achieving low volatility, but it can concentrate heavily in the single calmest holding.

Hierarchical ERC (HERC)

HERC uses the same clustering as HRP but allocates weight down the real cluster tree with an equal-risk split at each branch. In practice it spreads more evenly than HRP — a good middle ground when HRP over-concentrates in one low-volatility holding while you still want the clustering to be respected.

Diversification: the effective number of bets

A portfolio can look diversified (many holdings) while really being one or two bets in disguise (they all move together). The effective number of bets measures true diversification: it's high when risk is spread evenly across independent sources, and low when a few holdings dominate.

In the comparison table it's shown as X / N — the effective number out of your N holdings. Equal Risk Contribution pushes this toward the maximum; a concentrated allocation pushes it toward 1.

What's driving your returns (factor exposure)

The "What's driving your returns" panel breaks your portfolio's ups and downs into a handful of common market factors (or "themes") — and shows how much is left over that's unique to your specific mix.

What a factor really is. A factor is a theme that big investors — pension funds, endowments, hedge funds — allocate money to as a group: the overall stock market, small companies, value vs. growth, momentum, quality, interest rates, corporate credit. When those investors move money into or out of a theme, every asset tied to that theme tends to move together. That shared movement is why two holdings with different names can still rise and fall in sync — they're riding the same theme. Measuring your factor exposure tells you which of these themes you're really betting on, whether you meant to or not.

How to read it.

• Each bar is your portfolio's exposure to one theme. A bar to the right means your portfolio moves with that theme; a bar to the left means it moves opposite. A longer bar means a stronger tie.
• The number on each bar is the strength of that tie (its "beta" — how much your portfolio moves when the theme moves 1%). The small % next to it is how much of your portfolio's ups and downs that theme accounts for; those shares add up to "Explained."
• Explained (R²) is how much of your portfolio's movement these common themes account for. The rest is unique to your particular holdings.
• Faint grey bars are too small to be a dependable pattern, so don't read much into them.

The plain-English factors: U.S. stock market (how much you simply move with stocks overall), Small companies (a tilt toward smaller vs. larger firms), Value vs. growth (cheaper "value" stocks vs. pricier "growth" names like big tech), Interest-rate sensitivity (how much you move with long-term rates, via long-term government bonds), and Corporate credit risk (exposure to riskier corporate bonds, which tend to fall alongside stocks in a crisis). Turning on the extra options adds Momentum (recent winners), Quality (profitable, sturdy companies), and Low-volatility (calmer stocks).

How it's computed: your portfolio's monthly returns are compared against low-cost, tradeable ETF proxies for each theme (built as long-vs-short pairs so each theme is measured cleanly). It's a measurement of your past exposure — not a forecast — and exposures can change over time.

Comparing methods

The "Compare weighting methods" table (on an editable portfolio) shows every method side by side, so you can see the trade-off before choosing:

• Volatility — the resulting annualized volatility of the portfolio.
• Top holding — the single largest weight (concentration of capital).
• Top risk share — the single largest share of risk (concentration of risk).
• Diversification — the effective number of bets (higher is more balanced).

Returns aren't shown because these are risk-first methods — they decide weights from risk and correlation, not from a forecast of returns. Click Apply on any row to set your weights, then review the full results above.

---

Not investment advice

These tools are for education and analysis. All figures are computed from historical data, which does not predict future results. MarketHeist does not provide personalized investment advice.