Systematic Energy Trading

Systematic Energy Trading

Strategy Description

This essay explores trading opportunities on WTI futures in the energy market through 2 risk premia: moreover, carry-momentum and reaction-function-based momentum. 

Carry is a well-known strong systematic signal for WTI. It reflects cumulative imbalances between supply and demands (inventory levels), which is a principle component of oil pricing. However, this signal is slow as contango and backwardation don’t flip frequently, in which case, we won’t change our position frequently, and we may lose opportunities. As a result of trying to avoid such potential losses, and make more profit by quick update, we blend carry and momentum signals by applying momentum directly on carry (term-structure) to speed up our position switch. This new carry-momentum signal captures the instantaneous changes in supply and demand, thus more sensitive.

Momentum alone is also a strong signal that is explained by many theories.

Investors’ under-reaction to news, inertia in updating their beliefs, overconfidence, reluctance to change views, chasing, fear of missing out (FOMO), herding ignited by mutual reinforcement and social media-traders’ trading behavior, for example.

We then introduce the reaction function to map the momentum signals into positions. The reaction function describes the consideration: when momentum is too strong, the likelihood of reversals increases, and we will not invest too much in that situation. In this way, we improve the performance of momentum by sizing.

Since these 2 risk premia capture different factors that account for profit, we believe by combining them, we can achieve better result.

Portfolio Components

Underlying Asset, Investment Period & Adjustment

This portfolio consist of WTI futures(F1) only, and invest on the rollover-adjusted time series of daily prices, however, we back-tested performances of single and combined factors from Jan 1 2010 to Dec 31 2021, and test the resulting optimal portfolio’s performance from Jan 1 2022 to Feb 28 2022. 

Transaction costs are taken into consideration. Contracts are rolled 5 business days before expiry. 

Evaluation Metrics

Moreover, performances of factors are evaluated by for metrics: average annual P&L in dollar, annualized Sharpe Ratio, maximum drawdown, and return on drawdown(RoD). 

Furthermore, based on these metrics, parameters for the optimal portfolio are decided, it should strike a balance between risk and profit. 

Carry-Momentum

Basic Carry Strategy

By comparing 3 different carry (F1-F2, F1-F3, F1-F4) with epsilon = 0, we found that (F1-F3) serves as the best signal, i.e. n=3, that has the highest annualized P&L, Sharpe ratio, and lowest Maximum Drawdown.

Carry Momentum Strategy

Based on the basic carry signal (F1-F3), we compared different look-back periods for momentum(3 days, one week, half month, and one month). We found that when N =10 (half month), the carry-momentum signal has the highest annualized profit, the highest Sharpe ratio, the lowest maximum drawdown, and the highest RoD.

RF-Based Momentum

Basic Momentum Strategy

Momentum strategy performs best with n=20 as the look-back parameter for moving average. It has performed best out of 5 in all 4 metrics: $6212.60 annualized P&L, 0.2867 Sharpe Ratio, $43570.00 maximum drawdown, 0.1426 RoD.

Reaction-Function-Based Momentum

In reality, when momentum is too strong, the likelihood of reversals increases, and we will not invest too much in that situation. Therefore we introduced the reaction function. 

The adjusted strategy witnessed a great improvement with B=1. It has Sharpe Ratio of 0.4396, and RoD of 0.2521. Since the average position is different from the basic momentum strategy, it can’t be compared by annualized P&L or Maximum Drawdown, which depends on absolute values.

Optimal Combination

We combined two strategies since it describes different factors, in the hope of capturing more returns.

The optimal combination puts 73% of its weight on carry-momentum and 27% of it on reaction-function-based momentum. Thus, it achieved annualized portfolio of $16475.87, Sharpe Ratio of 0.9574, maximum drawdown of $24571.04, RoD of 0.6705. The relative performance metrics suggest it is better than either single-factor strategy.

Performance Test & Analysis

We applied our optimal strategy on WTI futures during the first two months of 2022. The portfolio yields an annualized P&L of $65020.52, Sharpe Ratio of 2.87, maximum drawdown of $6963.78, RoD of 9.34. 

It seems our strategy performs pretty well, but when compared with the equity line, we think it is largely due to the fact that overall oil market going up.

Under recent events, i.e. Russia’s invasion of Ukraine, and corresponding sanctions against Russia’s export of crude oil, moreover the global oil market experienced a surge in price. However, while our strategy smoothed peril under extreme situations, such as during pandemic, it also gives up probabilities of high yields, such as during wars.

Lastly, it is more applicable in mid- or long-term investment than event-driven speculation. 

Author: Jiaxuan Liang 

Back To News

Systematic Energy Trading