The advent of advanced language processing models like OpenAIs GPT-4 presents new opportunities for enhancing financial decision-making. This study aims to explore the potential of GPT-4 in forecasting the performance of quantitative trading strategies, with a focus on the application of specific indicators in a long-short portfolio over a time frame. To achieve this, we employ a novel approach that involves posing targeted questions to GPT regarding the effectiveness of the indicators. The responses of the model are then subjected to a comparison with backtesting results obtained from the corresponding timeframe, enabling an evaluation of the predictive accuracy of GPT-4. By leveraging the linguistic capabilities of GPT, we aim to extract predictions that can inform the optimization of strategies. The benchmarking results obtained from this comparison serve as the primary output of the study, offering an objective assessment of the model's performance in forecasting the behavior of markets.