Rosetta Analytics Inc. : RL Carbon

Year-to-Date
21.58%
Aug Performance
-0.11%
Min Investment
$ 1,000k
Mgmt. Fee
2.00%
Perf. Fee
25.00%
Annualized Vol
16.33%
Sharpe (RFR=1%)
1.66
CAROR
-
Assets
$ 6.2M
Worst DD
-7.60
S&P Correlation
0.07

Growth of 1,000 - VAMI

Monthly Performance

Export Data
Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec YTD DD

Past performance is not necessarily indicative of future results. The risk of loss in trading commodity futures, options, and foreign exchange ("forex") is substantial.

Period Returns

Program / Index Aug Qtr YTD 1yr 3yr 5yr 10yr Since
5/2020
RL Carbon -0.11 3.21 21.58 26.01 - - - 42.55
S&P 500 3.04 7.72 24.49 33.58 - - - 53.59
+/- S&P 500 -3.15 -4.51 -2.90 -7.58 - - - -11.04

Strategy Description

Summary

RL Carbon is based entirely on Rosetta’s deep reinforcement learning (DRL) model. Deep reinforcement learning is a type of machine learning that produces actions, not predictions, that are most likely to maximize (or minimize) a metric over an infinite time horizon. Beginning simply... Read More

Account & Fees

Type Managed Account
Minimum Investment $ 1,000k
Trading Level Incremental Increase $ 0k
CTA Max Funding Factor 3.00
Management Fee 2.00%
Performance Fee 25.00%
Average Commission $5.00
Available to US Investors Yes

Subscriptions

High Water Mark Yes
Subscription Frequency 1-7 Days
Redemption Frequency 1-7 Days
Investor Requirements QEP
Lock-up Period 0

Trading

Trading Frequency 470 RT/YR/$M
Avg. Margin-to-Equity 5%
Targeted Worst DD
Worst Peak-to-Trough
Sector Focus Global Climate Traders

Holding Periods

Over 12 Months 0%
4-12 Months 0%
1-3 Months 0%
1-30 Days 100.00%
Intraday 0%

Decision-Making

Discretionary 0%
Systematic 100.00%

Strategy

Other
100.00%
Strategy Pie Chart

Composition

Other
100.00%
Composition Pie Chart

Summary

RL Carbon is based entirely on Rosetta’s deep reinforcement learning (DRL) model. Deep reinforcement learning is a type of machine learning that produces actions, not predictions, that are most likely to maximize (or minimize) a metric over an infinite time horizon. Beginning simply with data (the model is not programmed), the model undertakes a series of random actions. Through a process of trial and error, it receives feedback from its environment, and over time it discovers sequences of actions that provide the best results. This iterative process makes DRL particularly good at solving dynamic optimization problems like allocating risk capital, which is why we chose to use DRL as the investment process for our RL Carbon strategy. In our case, our DRL model seeks to maximize the cumulative risk-adjusted return on capital invested EUA emission futures. Starting only with data and accounting for transaction costs, the model continually adjusts its behavior to learn the optimal daily allocation of risk capital to ICE EUA futures contracts to best achieve this return. These actions, i.e., investment decisions, are expressed as a single value (or signal) ranging from 100% long to 100% short. For example, a signal of 56% would require a 56% allocation of risk capital to long exposure to the EUAs and a 44% allocation to cash. We implement these actions by taking long or short positions in the front-year December EUA futures contracts.

Investment Strategy

RL Carbon uses Rosetta’s proprietary deep reinforcement learning model to determine the optimal size of long or short positions in the ICE EUA emissions futures contracts on a T+1 basis. It implements these predictions by taking unleveraged long or short positions in EUA futures contracts.

Risk Management

Risk management is embedded in the model: the deep reinforcement learning model considers risk and transaction costs when determining the optimal long or short allocation to EUA futures.

   

Past performance is not necessarily indicative of future results. The risk of loss in trading commodity futures, options, and foreign exchange ("forex") is substantial.

Reward
Average RoR:
Max Gain:
Gain Frequency:
Average Gain:
Gain Deviation:
Risk
Standard Deviation:
Worst Loss:
Loss Frequency:
Average Loss:
Loss Deviation:
Reward/Risk
Sharpe Ratio: (RF=1%)
Skewness:
Kurtosis:
Reward
Compound RoR:
Average RoR:
Max Gain:
Gain Frequency:
Average Gain:
Gain Deviation:
Risk
Standard Deviation:
Worst Loss:
Loss Frequency:
Average Loss:
Loss Deviation:
Reward/Risk
Sharpe Ratio: (RF=1%)
Skewness:
Kurtosis:

Past performance is not necessarily indicative of future results. The risk of loss in trading commodity futures, options, and foreign exchange ("forex") is substantial.

Note: Figures shown in the Monthly column are the greatest figures (or worst for losses/drawdowns) for any particular month. The Annual figures are the greatest for any calendar year.

Drawdown Report

Depth Length (Mos.) Recovery (Mos.) Peak Valley
-7.60 5 2 9/1/2020 2/1/2021
-2.77 2 - 6/1/2021 8/1/2021
-0.46 1 1 7/1/2020 8/1/2020
-0.05 1 1 1/1/0001 5/1/2020
Show More

Consecutive Gains

Run-up Length (Mos.) Start End
30.05 4 3/1/2021 6/1/2021
13.71 2 6/1/2020 7/1/2020
7.84 1 9/1/2020 9/1/2020
3.97 2 11/1/2020 12/1/2020
Show More

Consecutive Losses

Run-up Length (Mos.) Start End
-7.57 1 10/1/2020 10/1/2020
-3.85 2 1/1/2021 2/1/2021
-2.77 2 7/1/2021 8/1/2021
-0.46 1 8/1/2020 8/1/2020
-0.05 1 5/1/2020 5/1/2020
Show More

Top Performer Badges

Index Award Type Rank Performance Period

Past performance is not necessarily indicative of future results. The risk of loss in trading commodity futures, options, and foreign exchange ("forex") is substantial.