Risk Management in High-Volatility Assets

Abstract

Risk management begins by naming the risk being measured. Volatility, drawdown, tail loss, leverage and execution risk describe different problems. A strategy can improve one measure while leaving another unchanged—or making it worse. A useful evaluation keeps these distinctions visible instead of reducing them to one score.

1. Different measures answer different questions

Volatility describes dispersion in returns; it does not identify a worst possible loss. Drawdown measures a decline from a prior portfolio peak. The largest drawdown in a historical sample is an observation, not a limit on future losses.

Value-at-Risk is a quantile estimate for a stated horizon and confidence level. Expected Shortfall describes losses beyond the relevant threshold under the model or empirical sample. Both depend on the chosen data, frequency and assumptions. A parametric estimate and an empirical estimate should not be presented as interchangeable.

2. Why changing risk can motivate changing exposure

Moreira and Muir’s Volatility-Managed Portfolios reports benefits from reducing exposure when volatility is high across the asset-pricing factors and currency carry portfolios they study. That result motivates a question about risk budgeting; it does not validate Bitcoin timing or any Azimuth engine. Moreira & Muir, 2017 ↗

A proposed allocation rule still has to specify what it observes, when it changes exposure and how it pays for those changes. If its benefit disappears after plausible execution costs or a small parameter change, the apparent improvement deserves caution.

3. A useful limit is not a guarantee

An exposure cap limits an intended position, not the price at which it can be closed. A stop instruction depends on execution. A historical loss threshold describes the research sample. These are three different statements, and the language used to describe them should remain different.

The same distinction applies to fully funded investing. Avoiding borrowing removes one source of financing risk; it does not remove the possibility of losing the invested capital.

4. Include the open-position path

A realized-profit curve counts closed outcomes. A Mark-to-Market curve also values positions while they remain open. In a futures strategy, this difference can be material: a favorable exit does not erase the adverse path that preceded it.

A favorable completed trade does not erase the adverse path before exit. A strategy comparison should use consistently valued open positions, a disclosed observation frequency and documented execution assumptions. Closed-trade smoothness alone does not establish low portfolio risk. Read the evidence scope ↗

5. Stress the assumptions, not just the chart

A useful evaluation asks what happens with slower execution, higher costs, wider spreads, different start dates or a missing period of exceptional returns. It should also ask what the model cannot simulate. A scenario can make an assumption visible without assigning a reliable probability to that scenario.

Conclusion

The purpose of a risk framework is to make decisions and limitations explicit. It should describe which risks are constrained, which are estimated and which remain outside the model. None of these categories supports a promise of profit or a maximum future loss.