RR Sortino A Sharper Ratio
RR Sortino A Sharper Ratio
A ‘Sharper’ Ratio
Thomas N. Rollinger
Scott T. Hoffman
www.RedRockCapital.com
Sortino: A ‘Sharper’ Ratio | By Thomas N. Rollinger & Scott T. Hoffman | Red Rock Capital
Many traders and investment managers have the desire to measure and compare CTA
managers and / or trading systems. We believe risk-adjusted returns are one of the most
important measures to consider since, given the inherent / free leverage of the futures
markets, more return can always be earned by taking more risk. The most popular measure
of risk-adjusted performance is the Sharpe ratio. While the Sharpe ratio is definitely the most
widely used, it is not without its issues and limitations. We believe the Sortino ratio improves
on the Sharpe ratio in a few areas. The purpose of this article, however, is not necessarily to
extol the virtues of the Sortino ratio, but rather to review its definition and present how to
properly calculate it since we have often seen its calculation done incorrectly.
Sharpe Ratio
The Sharpe ratio is a metric which aims to GRAPH 2
measure the desirability of a risky investment Same Sharpe, much different risk.
strategy or instrument by dividing the average
period return in excess of the risk-free rate by mean mean
the standard deviation of the return generating typical CTA trend typical options
process. Devised in 1966 as a measure of per- following program seller program
SORTINO RATIO
(R – T)
S= Where:
TDD
Xi = ith return
N = total number of returns
u = average of all Xi returns.
The differences are: First, calculate the numerator of the Sortino ratio,
the average period return minus the target return:
1. In the Target Downside Deviation calcula-
tion, the deviations of Xi from the user select- Average Annual Return - Target Return:
able target return are measured, whereas in the (17% + 15% + 23% - 5% + 12% + 9% + 13% - 4%)
Standard Deviation calculation, the deviations ÷ 8 – 0% = 10%
of Xi from the average of all Xi is measured. Next, calculate the Target Downside Deviation:
2. In the Target Downside Deviation calcu- 1. For each data point, calculate the difference
lation, all Xi above the target return are set to between that data point and the target level. For
zero, but these zeros are still included in the summa- data points above the target level, set the differ-
tion. The calculation for Standard Deviation has ence to 0%. The result of this step is the under-
no Min() function. performance data set.
Standard deviation is a measure of dispersion 17% - 0% = 0%
of data around its mean, both above and below. 15% - 0% = 0%
Target Downside Deviation is a measure of dis- 23% - 0% = 0%
persion of data below some user selectable target -5% - 0% = -5%
return with all above target returns treated as 12% - 0% = 0%
underperformance of zero. Big difference. 9% - 0% = 0%
13% - 0% = 0%
Example Sortino -4% - 0% = -4%
Ratio Calculation 2. Next, calculate the square of each value in
the underperformance data set determined in
In this example, we will calculate the annual Sor- the first step. Note that percentages need to be
tino ratio for the following set of annual returns: expressed as decimal values before squaring, i.e.
Annual Returns: 5% = 0.05.
17%, 15%, 23%, -5%, 12%, 9%, 13%, -4% 0%^2= 0%
Target Return: 0% 0%^2 = 0%
Although in this example we use a target return 0%^2 = 0%
of 0%, any value may be selected, depending -5%^2 = 0.25%
on the application, i.e. a futures trading system 0%^2 = 0%
developer comparing different trading systems 0%^2 = 0%
vs. a pension fund manager with a mandate to 0%^2 = 0%
achieve 8% annual returns. Of course using a -4%^2 = 0.16%
different target return will result in a different
3. Then, calculate the average of all squared
value for the Target Downside Deviation. If you
differences determined in Step 2. Notice that
are using the Sortino ratio to compare managers
we do not “throw away” the 0% values.
or trading systems, you should be consistent in
using the same target return value. Average: (0% + 0% + 0% + 0.25% +
0% + 0% + 0% + 0.16%) ÷ 8 = 0.0513%
4. Then, take the square root of the average we see the target downside deviation calculated
determined in Step 3. This is the Target by “throwing away all the positive returns and
Downside Deviation used in denominator of take the standard deviation of negative returns”.
the Sortino Ratio. Again, percentages need We hope that by reading this article, you can see
to be expressed as decimal values before how this is incorrect. Specifically:
performing the square root function.
• In Step 1 above, the difference with respect
Target Downside Deviation: to the target level is calculated, unlike the
Square root of √0.0513% = 2.264% standard deviation calculation where the
difference is calculated with respect to the
Finally, we calculate the Sortino ratio: mean of all data points. If every data point
10% equals the mean, then the standard deviation
= 4.417 Sortino Ratio
2.264% is zero, no matter what the mean is. Con-
sider the following return stream: [-10, -10,
COMPARISON OF SORTINO RATIOS -10, -10]. The standard deviation is 0, while
the target downside deviation is 10 (assum-
WINTON 1.95
ing target return is 0).
RED ROCK 1.78
• In Step 3 above, all above target returns are in-
LYNX 1.59
cluded in the averaging calculation. The above
BLUETREND 1.56
target returns set to 0% in step 1 are kept.
TRANSTREND 1.42
• The Sortino ratio takes into account both
MAN AHL 1.22
the frequency of below target returns as well
NEWEDGE CTA 1.11
as the magnitude of below target returns.
ASPECT 0.87 Throwing away the zero underperformance
GRAHAM 0.74 data points removes the ratio’s sensitivity to
SUNRISE 0.55 frequency of underperformance. Consider
The above table summarizes the Sortino ratios for a popular the following underperformance return
CTA index and a handful of well-known CTAs (from the Sep- streams: [0, 0, 0, -10] and [-10, -10, -10,
tember 2003 inception of Red Rock Capital’s Systematic Global -10]. Throwing away the zero underperfor-
Macro program through March 2015). Graham is K4D-15V, mance data points results in the same target
Sunrise is Expanded Diversified, BlueTrend is from April 2004
downside deviation for both return streams,
and is now Systematica. Source: BarclayHedge. Past perfor-
mance is not necessarily indicative of future performance. but clearly the first return stream has much
less downside risk than the second.
Important Disclosures
This document is for informational purposes only and it is not a solicitation for investment. Past
results are not necessarily indicative of future results. An investment with any Commodity Trading
Advisor should only be made after careful study of the advisor’s Disclosure Document including the
description of the objectives, principal risks, charges, and fees associated with such an investment.
Red Rock Capital is an award-winning commodity investment management firm. During the sum-
mer of 2015 Red Rock will proudly celebrate its 12th anniversary. The firm is led by Thomas Roll-
inger, who is most notably a devoted pupil and former protégé of quantitative hedge fund legend,
Edward O. Thorp. Given recent developments with the firm, plus increasingly favorable market
conditions, Red Rock is especially well-positioned to continue to grow and thrive in the managed
futures industry.