What is being distributed
Distinguish issuer-specific risk from broad market risk. If one company loses an important customer, other businesses may be unaffected. A financing deterioration or broad downturn can affect many at once. Spreading investments does not promise to remove joint losses.[5](#source-save-risk)[9](#source-valuation-allocation)
Look through the wrapper. Several funds may own the same largest assets; two securities may belong to one group; different sector labels may depend on one commodity price or buyer group. This requires holdings and income-mechanism information, not a short name. Separate market concentration from operational concentration: different custodians may reduce dependence on one access system without changing the economics of identical assets, while varied assets in one account do not answer every access question.
A single-company risk example
All weights are fictional and are not a proposed allocation. If half the portfolio value is in one business and its value falls 40% while the rest is unchanged, the portfolio result is minus 20%.
If ten independent educational positions have equal starting weights and only one loses 40%, the portfolio falls 4%. This is the arithmetic limit of one position’s influence; it does not prove that ten real assets are independent or that the others remain unchanged. If all ten fall together, the number of lines does not save the result. Test both a single-participant problem and a common factor affecting the group.
Joint movement and correlation
Correlation describes joint changes over a chosen period; it is not a permanent property of two names. Results depend on data and observation window. Be especially careful assuming a weak prior-year relationship survives stress.
In a separate educational statistical model, two parts have equal weights and identical annual volatility of 20%. With correlation of one, volatility of their sum is also 20%; with zero correlation it is about 14.14%. These numbers come from the stated variance formula, not real-market observations. Statistical dispersion reduction does not remove permanent-loss or liquidity risk.
Checking hidden common factors
Create a dependency map: issuer and group, sector, geography, currency of future expenses, revenue source, financing method and holdings of underlying funds. The goal is not the maximum number of categories but visible, explained concentrations.
Review the map after fund changes, corporate events or a sharp rise in one position. Concentration can increase without a new purchase. If disclosure is insufficient, say independence was not checked; unknown holdings cannot be treated as diversified by default. Diversification is useful as a testable mechanism, not a promise of a positive result in every period, and it does not replace reserves, instrument analysis or a clear money-needed date.