Volatility Scaling: Keeping Portfolio Risk on Target

Most portfolios are built once and left to drift. An investor sets allocations to hit a target risk level — a classic 60/40 split, say — and assumes the risk profile will hold. It will not. Market volatility is not a constant. It clusters, spikes violently during crises, and compresses for long stretches in between. A portfolio calibrated to a comfortable risk band at inception can quietly double its realised volatility in a stress episode, precisely when the investor can least afford the surprise.

This is the problem that volatility scaling — also called volatility targeting or risk targeting — is designed to solve. It is one of the most widely used techniques in systematic asset management, and it sits at the centre of large institutional vehicles such as Man AHL's TargetRisk fund (roughly USD 2.6 billion under management). A recent empirical study covering December 2014 to January 2026 puts the approach under a microscope across eleven years and a remarkable sequence of shocks — a pandemic, a zero-rate era, the sharpest rate-hiking cycle in decades, and a sustained bull market. The findings are worth understanding for anyone responsible for keeping a portfolio inside a defined risk envelope.

What volatility scaling actually does

The mechanism is simpler than it sounds. You estimate the portfolio's current volatility from recent returns, then scale total exposure up or down so that realised risk tracks a fixed target — 10% annualised, for example. When markets are calm and measured volatility falls below target, the strategy gently levers up. When turbulence rises, it de-levers automatically. The result is a time-varying leverage factor that leans against the wind: more exposure in quiet markets, less in stormy ones.

Critically, this is a risk rule, not a return rule. The study finds the leverage factor is almost perfectly inversely correlated with realised volatility (a Spearman rank correlation of −0.99) yet effectively uncorrelated with next-day returns in equities, bonds or commodities. In plain terms: the strategy makes no bet on direction. It expresses no view on whether markets will rise or fall. It adjusts only the size of the risk, never its sign.

The headline result: risk becomes predictable

The single most robust finding is that volatility scaling delivers what it promises. Measured across estimation windows from 21 to 252 days, the dispersion of realised volatility — how far the actual risk level wanders from its target — falls by roughly 48 to 54% versus an unscaled portfolio. At longer measurement horizons the volatility profile becomes almost flat. Two formal statistical tests confirm the effect at the highest confidence levels.

For a family office or advisor, this is the practical payoff. A portfolio that genuinely holds its risk band is easier to match against liabilities, easier to govern against a mandate, and far easier to explain to a client. The number on the risk report means what it says, month after month.

The catch every investor should understand

Here is where the evidence becomes genuinely useful, because it is also a warning. Volatility scaling is not a free lunch, and it is emphatically not a crash hedge. Its performance is sharply regime-dependent.

In the calmest quarter of market conditions, the scaled strategy shines: it delivered a Sharpe ratio of 0.75, nearly double the 0.41 of a static 60/40 benchmark. This is the environment where the rule levers up and harvests more of the risk premium per unit of risk. But in the most stressed quarter, the relationship inverts — the static benchmark won handily (Sharpe 0.83 versus 0.38).

The reason is mechanical and important. Because the strategy de-levers as volatility spikes, it cuts exposure at the moment of maximum turbulence — which is often the market bottom. It then re-levers only slowly as calm returns. In a V-shaped recovery like March 2020, that timing is costly: the study shows the scaled strategy returned just +1.3% in 2020 against +16.6% for the 60/40 mix. The technique stabilises risk by surrendering some upside in violent rebounds. That is not a flaw — it is the explicit trade the investor is making, and it should be understood going in.

Robust, and cheap to run

Two further results matter for implementation. First, the benefit is robust to design choices: across eight different volatility-estimation windows, the scaled strategy beat the unscaled baseline on risk-adjusted return every time. Shorter windows react faster and time better but trade more; longer windows are smoother but slower. There is no single magic setting, which is reassuring — the effect does not rest on a fragile parameter choice.

Second, it is inexpensive to operate in liquid instruments. At institutional ETF trading costs of around 10 basis points, the drag was just 0.05 on the Sharpe ratio, and the strategy stayed worthwhile up to roughly 25–30 basis points of round-trip cost — well above what large investors actually pay. Daily rebalancing of a target-risk portfolio is not the cost problem some assume.

What it means for how we think about risk

The deeper lesson sits underneath the statistics. Volatility scaling reframes the central question of portfolio construction. The job is no longer to pick a fixed allocation and hope the risk behaves. It is to decide what level of risk you want to hold constant, and then let exposure flex to defend that level. For investors whose mandates, liability profiles or client expectations are calibrated to a specific volatility band, that is a meaningfully different — and more honest — way to run money.

It also clarifies what target-volatility strategies are for. They are not designed to dodge crashes; a rule that de-levers in response to volatility will always be a step behind a sudden shock. They are designed to keep the experience of holding the portfolio inside predictable bounds across the long middle of the cycle, where most of investing actually happens. Used with that understanding — as a discipline for stability rather than a hedge against disaster — the evidence for volatility scaling is strong and consistent.

At Vayo Capital, this is the lens we bring to multi-strategy portfolio construction: risk is the thing you control directly; return is what you earn for holding it steady.

---

This article draws on "Volatility Scaling in Multi-Asset Portfolios: Evidence from a Systematic Risk-Targeting Strategy" (de Souza e Almeida and Guedes de Farias, 2026), an empirical study of eight liquid ETFs spanning global equities, fixed income, inflation-linked bonds and commodities from December 2014 to January 2026. Figures cited illustrate the strategy's documented behaviour and are not a representation of Vayo Capital returns. Nothing here is investment advice.

Frequently asked questions

What is volatility scaling? Volatility scaling — also called volatility targeting or risk targeting — is a systematic technique that adjusts a portfolio's total exposure so its realised risk stays close to a fixed target. Exposure rises when markets are calm and falls when volatility spikes.

Does volatility targeting protect against market crashes? Not directly. Because it de-levers in response to rising volatility, it reduces exposure near the point of maximum turbulence — often the market bottom — and re-levers slowly. It stabilises the risk profile but can underperform static portfolios during sharp V-shaped recoveries.

When does volatility scaling work best? The evidence shows it adds the most value in calm market regimes, where it levers up and earns a higher risk-adjusted return than a static 60/40 portfolio. In severe stress episodes, static benchmarks tend to outperform.

Is volatility scaling expensive to run? In liquid instruments such as ETFs, no. At around 10 basis points of trading cost the performance drag is minimal, and the approach remains viable well above typical institutional execution costs.