Average returns describe where a programme ended up. Drawdowns describe what holding it felt like along the way. Allocators report losses to committees while those losses are happening, and the shape of a decline often decides whether an allocation is kept long enough to deliver its long-run result. This article defines the main drawdown and recovery measures, shows how they are calculated and lists the measurement choices that change them.

Defining a drawdown

A drawdown is the percentage decline in cumulative value from its highest level to date, the running peak, to a later value. For any month: drawdown = value ÷ running peak − 1. The series is zero whenever the programme sits at a new high and negative otherwise. Plotted through time it forms the underwater curve, a standard panel in managed-futures reporting.

Anatomy of a drawdown. Illustrative.Anatomy of a drawdownIllustrativePeakTroughRecoveryDrawdown from running peakDepth -18%Time to trough (14 m)Time to recovery (26 m)Synthetic monthly series. Not the record of any programme.
Figure 1. A drawdown on a synthetic monthly series. Upper panel: cumulative value with its running peak (dashed). Lower panel: the underwater curve. In this example the decline reaches 18% over 14 months and takes 26 months to regain the prior peak.

Three dimensions of a loss

Depth

Maximum drawdown is the largest peak-to-trough decline over the measurement period. It is the most quoted risk figure in managed futures, and CFTC Regulation 4.35 requires a CTA's disclosure document to state the worst peak-to-valley drawdown of the programme offered. [1] Its weakness is that it is a single observation. Two programmes with the same maximum drawdown may have reached it once or five times.

Time to trough

The number of months from the peak to the lowest point. A sharp, short decline and a slow erosion over two years may share a depth and still test an investor in different ways. In trend-following programmes, gradual declines commonly reflect range-bound markets, while abrupt ones commonly follow sudden reversals of established trends.

Time to recovery

The months from the trough until cumulative value regains the prior peak. The whole episode, peak to recovery, is the period in which an investor who entered at the peak shows a loss. A drawdown that has not yet recovered is open; report its current depth and age alongside closed episodes, because leaving it out flatters the record.

The arithmetic of recovery

Gains and losses do not offset symmetrically. After a decline of d, the gain needed to return to the prior peak is d ÷ (1 − d). A 10% drawdown needs an 11.1% gain. A 20% drawdown needs 25%. A 50% drawdown needs 100%. The curve steepens as losses deepen, which is why allocators set drawdown limits and review triggers well before recovery would require an implausible run of returns.

Gain required to recover a drawdown. Arithmetic.Gain required to recover a drawdownArithmeticDrawdownGain needed to return to the prior peak−10%+11.1%−20%+25.0%−30%+42.9%−40%+66.7%−50%+100.0%Required gain = d ÷ (1 − d). Excludes fees, cash flows and interest.
Figure 2. Gain required to return to a prior peak after drawdowns of 10% to 50%. Calculated, not drawn from manager data.

Ratios built on drawdowns

Several ratios relate return to drawdown instead of volatility:

  • Calmar ratio. Annualised return divided by maximum drawdown, originally over a trailing 36 months. Terry Young introduced it in 1991 and named it after his California Managed Accounts Reports newsletter. [2]
  • Sterling ratio. Return divided by an average of annual maximum drawdowns, which reduces dependence on one episode. Definitions vary between data providers, including whether a fixed percentage is added to the denominator.
  • Ulcer index. Developed by Peter Martin in 1987, it takes the root mean square of percentage drawdowns over a period, so it reflects depth and duration together. [3]
  • Average drawdown and drawdown frequency. Descriptive measures that put the maximum in context.

Each of these depends on the window. A Calmar ratio over a period that excludes a programme's worst episode will differ sharply from one that includes it. Ask for the window, and compare like with like.

Measurement choices that change the answer

  • Frequency. Monthly data misses declines that opened and closed within a month. Daily account data, where a managed account provides it, shows the fuller picture.
  • Net or gross. Fees deepen and lengthen drawdowns. Use net figures on the terms on offer.
  • Hypothetical months. Simulated periods carry no execution friction. Report their drawdowns separately.
  • Funding level. For a partially funded account, a drawdown measured against cash deposited is larger than one measured against nominal account size. See Understanding CTA Managed Accounts.
  • Leverage. A programme offered at several risk levels shows proportionally different drawdowns. Compare at a common volatility where possible.

Using drawdown measures in allocation work

At the programme level, drawdown history informs sizing. An allocator who knows the depth and duration of past losses sets position limits and review points before capital is committed, not in the middle of a decline. At the portfolio level the useful question is overlap: two managers with moderate drawdowns that fall in the same months add less diversification than their all-month correlation suggests.

A current drawdown also needs context. Allocator IQ shows a programme's present drawdown depth against its own history, and the speed of its past recoveries, beside the standard measures across its manager universe. Neither is a forecast. Past recovery times describe what happened under conditions that may not repeat. For the wider review these measures sit in, see How to Evaluate a CTA Track Record; for comparison tools, see the platform.

Frequently asked questions

What is maximum drawdown?

Maximum drawdown is the largest percentage decline from a running peak in cumulative value to a later low over a stated period. CFTC rules require registered CTAs to disclose the worst peak-to-valley drawdown of the programme offered.

What is a good Calmar ratio?

No universal threshold applies. The ratio depends on the measurement window, the return frequency and whether the worst drawdown falls inside the window, so it is most informative when programmes are compared over the same period on the same basis.

How is time to recovery measured?

Time to recovery counts the months from the trough of a drawdown until cumulative value regains the prior peak. A drawdown that has not regained its peak is open, and its current depth and age should be reported.

Why do daily and monthly drawdowns differ?

Monthly data records only month-end values, so declines that open and recover within a month do not appear. Daily data captures those moves and gives a maximum drawdown at least as deep as the monthly figure, and often deeper.

For institutional investors. This article provides general information, not investment, tax or legal advice or a recommendation of any manager, strategy or structure. Futures trading involves substantial risk of loss and is not suitable for all investors. Past performance is not necessarily indicative of future results.

Sources

  1. CFTC Regulations, 17 CFR 4.35. Performance disclosures, including worst peak-to-valley drawdown.
  2. Young, T. W. (1991). Calmar Ratio: A Smoother Tool. Futures magazine.
  3. Martin, P. G. and McCann, B. B. (1989). The Investor's Guide to Fidelity Funds. Definition of the Ulcer index, first described by Martin in 1987.