Foundations4 min read

Asset Allocation

The split between asset classes is the decision that determines most of a portfolio's variability. It is also the one made least often and reviewed least carefully.

MadStockAlerts Research · Updated August 29, 2026

What to take away

  • Allocation across classes explains most of the variability of a portfolio's returns.
  • The classes behave differently because they respond to different things.
  • Correlations are estimates from a window and are unstable in stress.
  • An allocation drifts on its own as the components perform differently.
  • There is no allocation that is correct in general, only one that matches a set of constraints.

MAD Academy Training Video · 0:45

The Decision That Outweighs the Rest

How much sits in each asset class explains most of a portfolio's variability — far more than which specific securities were chosen.

This lesson is part of a Stock Alerts + Tools plan.

See the library

Why the split dominates

A widely cited body of research finds that the allocation across asset classes explains the large majority of the variability of a portfolio's returns over time, with security selection and timing accounting for far less. The finding is frequently overstated and the direction of it is robust.

The precise claim matters. The research addresses variability over time within a portfolio, not the difference in return between two portfolios, and those are different questions. The commonly repeated version conflates them.

What survives the caveats is still substantial: a portfolio that is eighty percent equities and one that is twenty percent equities will have different experiences in any given year, and the difference between them will swamp the effect of which particular equities were chosen.

What the classes are for

ClassResponds toRole usually assigned
EquitiesGrowth, earnings, discount ratesThe return engine, and the source of most variability
Government bondsRates and inflation expectationsIncome, and a diversifier in growth shocks
CreditRates plus default riskHigher income, and equity-like in stress
CashThe policy rateLiquidity and optionality
Real assetsInflation and supplyAn inflation hedge with substantial variance

The third row is the one that behaves differently from how it is often categorised. Corporate credit is frequently grouped with bonds and behaves substantially like equities in a downturn, which is when the grouping matters.

What each class is being held for
CashLiquidity and optionality. Tracks the policy rate
Government bondsIncome, and a diversifier in growth shocks
CreditHigher income, and equity-like when it matters
EquitiesThe return engine and most of the variability
Real assetsAn inflation hedge, with substantial variance
StabilityGrowth
Credit is the row that behaves differently from how it is usually filed. It sits with bonds on a statement and with equities in a downturn, which is when the classification matters.

Correlation is an estimate

Diversification depends on components not moving together, and how much they move together is measured over a historical window. The measurement is an estimate with error, and it is not stable.

  • Correlations rise in crises, which is exactly when the diversification was being relied upon.
  • The equity-bond correlation has changed sign across decades, and the sign depends on whether inflation or growth is the dominant shock.
  • A long window includes regimes that no longer apply; a short one is dominated by noise.
  • Assets can share an exposure without a historical correlation showing it, until the exposure is tested.

The second item is the most consequential in practice. A portfolio built on the assumption that bonds rise when equities fall is relying on a relationship that held in some decades and reversed in others.

Why there is no general answer

An allocation is a function of constraints rather than a function of markets. Two people facing identical markets can hold entirely different portfolios without either being wrong.

  • The horizon over which the money is needed, which determines how much variability is survivable.
  • Whether contributions continue, since a portfolio being added to behaves differently from one being drawn down.
  • Other income and its stability, since a secure income is itself a bond-like asset.
  • The account structure, which changes what is efficient to hold where.
  • The level of decline that would cause the plan to be abandoned, which is the binding constraint for most people.

The last is the one that is hardest to estimate and easiest to overestimate. An allocation that will be abandoned in a drawdown is worse than a more modest one that will be held, because abandonment happens at the bottom.

The allocation you can hold

Every allocation framework produces a recommendation, and the recommendation is only as good as the probability it is still in place after a bad year. That constraint is rarely modelled and frequently binding.

Equity shareApproximate decline in a severe equity bear marketWhat it requires
100%Around half the portfolioHolding through a loss of that size
80%Around 40 percentThe same, slightly reduced
60%Around 30 percentHistorically the most commonly held mix
40%Around 20 percentA materially different experience

The figures assume the fixed income portion holds up, which it did in several historical bear markets and did not in 2022. That is the correlation caveat made concrete rather than theoretical.

The question the table poses is not which row produces the best long-run return. It is which row would still be in place at the bottom, because an allocation abandoned in a decline realises the loss and misses the recovery.

Educational content only. MadStockAlerts provides market commentary, research, and educational content. It is not personalized investment advice, and nothing here is a recommendation to buy or sell any security. Trading and investing involve substantial risk, including loss of capital. See the Risk Disclosure and Customer Agreement.