Financial distress occurs when individuals or companies struggle to meet financial obligations, often due to factors like high fixed costs, illiquidity, or revenue sensitivity to economic downturns. While various machine learning and deep learning approaches have been developed for predicting financial distress by analyzing financial data to assess the likelihood of a company facing financial difficulties, challenges persist in achieving enhanced prediction accuracy and managing time complexity. To address this, a novel method known as the Chinese Whisper Clustered Stochastic Gradient Descent Federated Learning method (CWCSGDFL) is introduced. It aims to improve the efficiency of financial distress prediction by grouping data samples using the Hamann Similarity Indexed Chinese Whispers clustering process, validating them with a matching coefficient, and addressing data imbalance. Also, Kaiser-Meyer-Olkin correlative targeted projection model is employed to choose the pertinent features for prediction. By achieving balanced clustering results with these selected features, CWCSGDFL reduces time complexity in financial distress prediction. Stochastic Gradient Gaussian Kernel Federated Learning approach is employed to preserve privacy while predicting distress, aided by kriging regression for model analysis. Stochastic gradient descent momentum optimizes the global objective function, ensuring accurate classification with minimal loss. Experimental results demonstrate that CWCSGDFL achieves an average accuracy of 96%, precision of 94%, recall of 98%, F-measure of 98%, and a time complexity of 36 ms, outperforming the existing techniques.